Advertisement

Low Rates of Gastrointestinal and Non-Gastrointestinal Complications for Screening or Surveillance Colonoscopies in a Population-Based Study

  • Author Footnotes
    ∗ Authors share co-first authorship.
    Louise Wang
    Footnotes
    ∗ Authors share co-first authorship.
    Affiliations
    Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania
    Search for articles by this author
  • Ajitha Mannalithara
    Affiliations
    Division of Gastroenterology and Hepatology, Stanford University School of Medicine, Stanford, California
    Search for articles by this author
  • Gurkirpal Singh
    Affiliations
    Division of Gastroenterology and Hepatology, Stanford University School of Medicine, Stanford, California

    Institute of Clinical Outcomes Research and Education, Woodside, California
    Search for articles by this author
  • Author Footnotes
    ∗ Authors share co-first authorship.
    Uri Ladabaum
    Correspondence
    Reprint requests Address requests for reprints to: Uri Ladabaum, MD, MS, GI/Hepatology, Stanford University School of Medicine, M211, 300 Pasteur Drive, Stanford, California 94305. fax: (650) 723-5488.
    Footnotes
    ∗ Authors share co-first authorship.
    Affiliations
    Division of Gastroenterology and Hepatology, Stanford University School of Medicine, Stanford, California
    Search for articles by this author
  • Author Footnotes
    ∗ Authors share co-first authorship.
Published:October 11, 2017DOI:https://doi.org/10.1053/j.gastro.2017.10.006

      Background

      The full spectrum of serious non-gastrointestinal post-colonoscopy complications has not been well characterized. We analyzed rates of and factors associated with adverse post-colonoscopy gastrointestinal (GI) and non-gastrointestinal events (cardiovascular, pulmonary, or infectious) attributable to screening or surveillance colonoscopy (S-colo) and non-screening or non-surveillance colonoscopy (NS-colo).

      Methods

      We performed a population-based study of colonoscopy complications using databases from California hospital-owned and nonhospital-owned ambulatory facilities, emergency departments, and hospitals from January 1, 2005 through December 31, 2011. We identified patients who underwent S-colo (1.58 million), NS-colo (1.22 million), or low-risk comparator procedures (joint injection, aspiration, lithotripsy; arthroscopy, carpal tunnel; or cataract; 2.02 million) in California’s Ambulatory Services Databases. We identified patients who developed adverse events within 30 days, and factors associated with these events, through patient-level linkage to California’s Emergency Department and Inpatient Databases.

      Results

      After S-colo, the numbers of lower GI bleeding, perforation, myocardial infarction, and ischemic stroke per 10,000-persons were 5.3 (95% confidence interval [CI], 4.8–5.9), 2.9 (95% CI, 2.5–3.3), 2.5 (95% CI, 2.1–2.9), and 4.7 (95% CI, 4.1–5.2) without biopsy or intervention; with biopsy or intervention, numbers per 10,000-persons were 36.4 (95% CI, 35.1–37.6), 6.3 (95% CI, 5.8–6.8), 4.2 (95% CI, 3.8–4.7), and 9.1 (95% CI, 8.5–9.7). Rates of dysrhythmia were higher. After NS-colo, event rates were substantially higher. Most serious complications led to hospitalization, and most GI complications occurred within 14 days of colonoscopy. Ranges of adjusted odds ratios for serious GI complications, myocardial infarction, ischemic stroke, and serious pulmonary events after S-colo vs comparator procedures were 2.18 (95% CI, 2.02–2.36) to 5.13 (95% CI, 4.81–5.47), 0.67 (95% CI, 0.56–0.81) to 0.99 (95% CI, 0.83–1.19), 0.66 (95% CI, 0.59–0.75) to 1.13 (95% CI, 0.99–1.29), and 0.64 (95% CI, 0.61–0.68) to 1.05 (95% CI, 0.98–1.11). Biopsy or intervention, comorbidity, black race, low income, public insurance, and NS-colo were associated with post-colonoscopy adverse events.

      Conclusions

      In a population-based study in California, we found that following S-colo, rates of serious GI adverse events were low but clinically relevant, and that rates of myocardial infarction, stroke, and serious pulmonary events were no higher than after low-risk comparator procedures. Rates of myocardial infarction are similar to, but rates of stroke are higher than, those reported for the general population.

      Graphical abstract

      Keywords

      Abbreviations used in this paper:

      CI (confidence interval), CPT (Current Procedural Terminology), CRC (colorectal cancer), EGD (esophagogastroduodenoscopy), GI (gastrointestinal), ICD-9 (Classification of Disease), MI (myocardial infarction), NS-colo (non-screening or non-surveillance colonoscopy), S-colo (screening or surveillance colonoscopy), SASD (Surgery and Services Databases), SEDD (State Emergency Department Databases), SID (State Inpatient Databases), USPSTF (United States Preventive Services Task Force)
      See editorial on page 473.

       Background and Context

      The full spectrum of serious non-gastrointestinal adverse events after screening colonoscopy has not been well characterized, and the absolute risks are uncertain because most previous studies lack non-colonoscopy controls.

       New Findings

      After screening or surveillance colonoscopy, the rates of serious non-gastrointestinal adverse events, including myocardial infarction, stroke, and serious pulmonary events, are low and no higher than after low-risk comparator procedures.

       Limitations

      A control group with no exposure to colonoscopy or another ambulatory procedure was not available for direct comparisons.

       Impact

      Quality assurance efforts relating to colonoscopy safety should focus on lower gastrointestinal bleeding and perforation, and not on non-gastrointestinal events.
      Colonoscopy is the most commonly performed gastrointestinal (GI) endoscopic test in the United States.
      • Peery A.F.
      • Dellon E.S.
      • Lund J.
      • et al.
      Burden of gastrointestinal disease in the United States: 2012 update.
      • Seeff L.C.
      • Richards T.B.
      • Shapiro J.A.
      • et al.
      How many endoscopies are performed for colorectal cancer screening? Results from CDC’s survey of endoscopic capacity.
      It is the test of choice to evaluate lower GI symptoms in multiple clinical scenarios, and it is a central component of all colorectal cancer (CRC) control efforts, whether as a primary screen or as follow-up after a less invasive screen.
      While colonoscopy is generally safe, its risks span a spectrum of GI and non-GI complications. The Canadian Task Force on Preventive Health Care recently recommended not using colonoscopy as a primary CRC screening test, in part because of its potential for harm.
      • Bacchus C.M.
      • Dunfield L.
      • et al.
      Canadian Task Force on Preventive Health Care
      Recommendations on screening for colorectal cancer in primary care.
      In a modeling project commissioned by the United States Preventive Services Task Force (USPSTF) to inform its updated recommendation on CRC screening, the volume of colonoscopies generated by a particular screening strategy was used as a proxy for harms and costs.
      • Zauber A.G.
      • Knudsen A.B.
      • Rutter C.M.
      • et al.
      Evaluating the benefits and harms of colorectal cancer screening strategies: a collaborative modeling approach.
      An accompanying systematic review identified 55 studies addressing the adverse effects of colonoscopy, but despite this rich literature, important gaps in knowledge were highlighted.
      • Lin J.
      • Piper M.
      • Perdue L.
      • et al.
      Screening for colorectal cancer: an updated systematic review for the USPSTF.
      Pooled estimates could be derived for the risk of perforation (4 [95% confidence interval (CI), 2–5] per 10,000) and major hemorrhage (8 [95% CI, 5–14] per 10,000) associated with screening colonoscopy, but few studies consistently defined or reported other serious harms including non-GI complications, and few included nonendoscopy comparator arms to estimate the fraction of adverse events attributable to colonoscopy itself, or contrasted diagnostic colonoscopy vs colonoscopy with biopsy or polypectomy.
      The recent linkage by patient-level identifiers of California’s State Ambulatory Surgery and Services Databases (SASD), State Emergency Department Databases (SEDD) and State Inpatient Databases (SID) allowed us to perform a large state-wide, population-based study of colonoscopy complications. We examined GI and non-GI complications in detail after screening/surveillance colonoscopy (S-Colo) vs non-screening/surveillance colonoscopy (NS-Colo) with or without biopsy, polypectomy, or intervention; we included 3 non-colonoscopy comparator groups presumed to be at low risk of systemic complications; and we explored the time-course and predictors of colonoscopy complications. Our study’s large sample size, diverse population, and unique design allowed us to address key knowledge gaps regarding the risks of colonoscopy.

      Methods

       Overview

      Our analysis addressed 30-day GI and non-GI complication rates in patients undergoing outpatient colonoscopy. Event rates after S-Colo or NS-Colo were compared with those in 3 comparator groups consisting of patients who underwent other common and minimally invasive procedures that are not expected to have major systemic complications, and that were selected as described below: joint injection/aspiration/lithotripsy, arthroscopy/carpal tunnel surgery, or cataract surgery. Adjusted comparisons accounted for demographic factors and comorbidities in an attempt to account for any differences in the types of patients undergoing each type of procedure.

       Data Sources and Study Population

      SASD, SEDD, and SID were developed for the Healthcare Cost and Utilization Project and include hospital-owned and nonhospital-owned ambulatory facilities for the state of California spanning from January 1, 2005 through December 31, 2011. Database descriptions, documentation, references, and quality control procedures are available.

      Agency for Healthcare Research and Quality. HCUP quality control procedures. 2017. Available at: https://www.hcup-us.ahrq.gov/db/quality.pdf. (Accessed May 15, 2017).

      Agency for Healthcare Research and Quality. Introduction to the HCUP State Ambulatory Surgery and Services Databases (SASD). 2017. Available at: https://www.hcup-us.ahrq.gov/db/state/sasddist/Introduction_to_SASD.pdf. (Accessed May 15, 2017).

      Agency for Healthcare Research and Quality. Introduction to the HCUP State Emergency Department Databases (SEDD). 2017. Available at: https://www.hcup-us.ahrq.gov/db/state/sedddist/Introduction_to_SEDD.pdf. (Accessed May 15, 2017).

      Agency for Healthcare Research and Quality. Introduction to the HCUP State Inpatient Databases (SID). 2017. Available at: https://www.hcup-us.ahrq.gov/db/state/siddist/Introduction_to_SID.pdf. (Accessed May 15, 2017).

      Agency for Healthcare Research and Quality. HCUP supplemental variables for revisit analyses. Available at: https://hcup-us.ahrq.gov/toolssoftware/revisit/revisit.jsp. (Accessed May 15, 2017).

      The current analysis was made possible by the recent assignment of unique patient-level identifiers in these databases, which created the novel capability of linking the databases at the level of individual persons.

      Agency for Healthcare Research and Quality. HCUP supplemental variables for revisit analyses. Available at: https://hcup-us.ahrq.gov/toolssoftware/revisit/revisit.jsp. (Accessed May 15, 2017).

      We identified persons who underwent colonoscopy or 1 of the comparator group procedures in an ambulatory setting in the SASD, and searched for complications occurring within 30 days, defined as presenting to an emergency department (SEDD) or inpatient setting (SID). Patient records on dates prior to the index procedures were used to determine pre-existing comorbidities as described below.
      Current Procedural Terminology (CPT) codes were used to identify procedures, informed by the Clinical Classifications Software for Services and Procedures. International Classification of Disease (ICD-9) codes were used to identify indications, complications, and comorbid diagnoses (see below). Screening colonoscopy codes (G-codes) are not included in SASD.

       Defining the Colonoscopy Cohorts

      CPT codes were used to identify persons who underwent colonoscopy and no comparator group procedures within ±30 days (Supplementary Table 1). Colonoscopies were first classified as S-Colo or NS-Colo using ICD-9-CM codes and clinical Classifications Software categories, as we described in detail previously.
      • Ladabaum U.
      • Levin Z.
      • Mannalithara A.
      • et al.
      Colorectal testing utilization and payments in a large cohort of commercially insured US adults.
      Colonoscopies were further subclassified as either diagnostic or with biopsy/intervention including polypectomy (Supplementary Table 1), yielding 4 subgroups:
      • 1)
        Screening/surveillance colonoscopy, diagnostic (S-Colo-Diag).
      • 2)
        Screening/surveillance colonoscopy, with biopsy/intervention (S-Colo-Int).
      • 3)
        Non-screening/surveillance colonoscopy, diagnostic (NS-Colo-Diag).
      • 4)
        Non-screening/surveillance colonoscopy, with biopsy/intervention (NS-Colo-Int).
      For persons who underwent multiple colonoscopies, the first colonoscopy encounter was used. In multivariable analyses, we controlled for colonoscopy as part of “double procedures” (esophagogastroduodenoscopy [EGD], esophagoscopy, sigmoidoscopy, or small bowel endoscopy codes on the same day) or within ±30 days of these other endoscopic procedures (Supplementary Table 2). In a sensitivity analysis, we considered only colonoscopies that were not on the same day or within ±30 days of these other procedures.

       Comparison Group Procedure Cohorts

      The ideal comparator group to distinguish what fraction of the post-colonoscopy adverse events can be attributed to colonoscopy vs the expected background rate in the population would consist of patients comparable to those who underwent colonoscopy, but who did not undergo colonoscopy. Because the colonoscopy cohort was derived based on exposure to colonoscopy as documented in the SASD, we could not construct a control group without exposure to any ambulatory procedure at all. The best available strategy was to construct comparator groups of persons represented in SASD who underwent other procedures. We reasoned that the best comparators would be common minor procedures expected to have low overall complication rates, which might approximate expected background rates in the population, and to undertake adjusted comparisons between groups. We supplemented this with comparisons to published background event rates in the general population.
      To identify appropriate comparator groups, a list of the top 30 CPT codes ranked by frequency was generated. Comparator groups were chosen by consensus among our research group based on their presumed low risk for systemic complications based on our clinical judgment, absence of other associated procedures in SASD that could confound analyses, and generalizability to large patient populations. Three comparator groups were selected a priori based on these criteria: 1) joint injection/joint aspiration/lithotripsy to represent very minimally or non-invasive minor procedures that usually do not require anesthesia and that were judged to have minimal systemic risk; 2) arthroscopy/carpal tunnel release to represent minor procedures generally performed with anesthesia; and 3) cataract operations representing minimally invasive procedures, admittedly in an older population. Individual sub-analyses for the joint injection, aspiration, and lithotripsy subgroups were performed to gain further insight into their higher complication rates compared with colonoscopy.
      Demographic factors, including age, race, gender, patient location, and comorbidities were pre-specified as covariates for adjusted comparisons. Complete CPT codes of each procedure were collected to identify persons who underwent any of the established comparator procedures without colonoscopy, EGD, sigmoidoscopy, or small bowel endoscopy codes within ±30 days (Supplementary Table 3).

       Complications and Their Time-course

      For the primary analyses, we defined complications as emergency department visits or hospital admissions within 30 days following the initial procedure, based on the standard 30-day period in the post-surgical literature. In secondary analyses, we explored what percentage of these complications led to hospitalization, as a marker of severity, and what percentage of the 30-day complications occurred within 7, 14, or 21 days of colonoscopy. This time-course analysis was performed to gain insight into the post-colonoscopy observation period that is relevant to capture post-procedure complications.
      Complications were classified into categories based on ICD-9-CM codes associated with the encounter through the clinical Classifications Software, cross-checked against published literature of both colonoscopy
      • Ko C.W.
      • Dominitz J.A.
      Complications of colonoscopy: magnitude and management.
      • Arora G.
      • Mannalithara A.
      • Singh G.
      • et al.
      Risk of perforation from a colonoscopy in adults: a large population-based study.
      and other common procedures (Supplementary Table 4).
      • Sporn E.
      • Petroski G.F.
      • Mancini G.J.
      • et al.
      Laparoscopic appendectomy–is it worth the cost? Trend analysis in the US from 2000 to 2005.
      • Kreder H.J.
      • Deyo R.A.
      • Koepsell T.
      • et al.
      Relationship between the volume of total hip replacements performed by providers and the rates of postoperative complications in the state of Washington.
      The categories were the following: GI (serious: perforation, lower GI bleeding, upper GI bleeding, colonic diverticulitis, small intestinal diverticulitis; non-serious: abdominal pain, nausea, noninfectious gastroenteritis); cardiac (acute myocardial infarction [MI], coronary atherosclerosis, congestive heart failure, cardiac dysrhythmia, structural heart disease); cerebrovascular (ischemic stroke, hemorrhagic stroke, transient ischemic attack); pulmonary (pneumonia, bronchitis); and infection (bacteremia, fever). We analyzed 30-day in-hospital and emergency department mortality as coded by discharge disposition in the SID and SEDD, but we caution that this is not a comprehensive ascertainment of mortality outside of these facilities.

       Comorbidities

      ICD-9-CM codes were used to define pre-existing comorbidities for each study subject based on Charlson Deyo comorbidities
      • Deyo R.A.
      • Cherkin D.C.
      • Ciol M.A.
      Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases.
      (Supplementary Table 5). Records in all 3 databases up to 365 days before the index procedure were searched, and a comorbidity index based on Charlson Deyo scoring was calculated, as previously described (Supplementary Table 5).
      • Deyo R.A.
      • Cherkin D.C.
      • Ciol M.A.
      Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases.

       Statistical Analysis

      We performed analyses using statistical analysis software SAS version 9.4 (SAS Institute Inc, Cary, NC). First, for each procedure group, raw complication rates and 95% CI were calculated without stratification or adjustment. Adjusted odds ratios and 95% CI were then calculated to compare complication rates between groups while accounting for potential baseline differences in the types of patients undergoing each type or procedure.
      Multivariate logistic regression analyses were performed to identify potential predictors of colonoscopy complications. We considered the following predictors a priori: age, gender, race/ethnicity, Charlson-Deyo comorbidity score, number of chronic conditions, patient location (urban, large rural town, small rural town, or isolated rural town), median household income quartiles as defined by zip code, insurance, NS-Colo, biopsy/intervention, and sigmoidoscopy, EGD, or small bowel endoscopy within ±30 days of colonoscopy. A subanalysis was completed after eliminating colonoscopies that were within ±30 days of sigmoidoscopy, EGD, or small bowel endoscopy. Statistical significance was set at a 2-sided alpha ≤0.05. We report the C-statistic for these models as a measure of how well the predictors account for the outcomes.

      Results

       Demographics: Colonoscopy Groups

      S-Colo was performed on 1,580,318 patients and included biopsy/intervention in 59.2%, and NS-Colo was performed on 1,222,070 patients and included biopsy/intervention in 39.7% (Tables 1 and 2). The majority of S-Colo patients were between 50–69 years of age and few had documented comorbidities; NS-Colo patients tended to be younger and more had documented comorbidities (Table 1).
      Table 1Characteristics of Colonoscopy and Comparison Groups
      CharacteristicColonoscopy screening/surveillance (%) (n = 1,580,318)Colonoscopy for other than screening/surveillance (%) (n = 1,222,070)Joint injections, facets, lithotripsy (%) (n = 324,114)Arthroscopies, carpal tunnel Surgery (%) (n = 853,170)Cataract surgery (%) (n = 838,266)
      Race
       White1,042,493 (66.0)739,335 (60.5)227,547 (70.2)546,610 (64.1)511,106 (61.0)
       Black55,894 (3.5)47,715 (3.9)13,305 (4.1)36,699 (4.3)34,772 (4.1)
       Hispanic180,491 (11.4)197,939 (16.2)43,209 (13.3)137,572 (16.1)134,056 (16.0)
       Asian or Pacific Islander148,372 (9.4)104,274 (8.5)14,417 (4.4)27,661 (3.2)97,899 (11.7)
       Native American1467 (0.1)1370 (0.1)371 (0.1)915 (0.1)742 (0.1)
       Other41,666 (2.6)29,707 (2.4)5996 (1.8)13,448 (1.6)19,830 (2.4)
       Missing109,935 (7.0)101,730 (8.3)19,269 (5.9)90,265 (10.6)39,861 (4.8)
      Gender
       Men768,858 (48.7)490,826 (40.2)149,004 (46.0)426,214 (50.0)346,797 (41.4)
       Missing17,640 (1.1)27,391 (2.2)4857 (1.5)35,546 (4.2)6754 (0.8)
      Age
       Average60.4 (10.7)57.7 (14.8)56.7 (16.0)50.2 (15.6)71.9 (10.9)
       <50141,512 (9.0)324,028 (26.5)109,588 (33.8)392,677 (46.0)28,877 (3.4)
       50–59674,295 (42.7)353,025 (28.9)75,600 (23.3)224,461 (26.3)75,433 (9.0)
       60–69437,237 (27.7)264,643 (21.7)61,551 (19.0)135,896 (15.9)194,473 (23.2)
       70–79247,945 (15.7)190,083 (15.6)48,448 (14.9)67,633 (7.9)324,119 (38.7)
       ≥8076,723 (4.9)85,266 (7.0)27,914 (8.6)23,681 (2.8)214,174 (25.5)
       Missing2606 (0.2)5025 (0.4)1013 (0.3)8822 (1.0)1190 (0.1)
      Charlson Deyo score
       Average0.2 (0.8)0.3 (1.0)0.3 (0.9)0.2 (0.6)0.6 (1.3)
       01,378,174 (87.2)1,010,591 (82.7)269,475 (83.1)734,332 (86.1)586,333 (69.9)
       1124,849 (7.9)118,594 (9.7)34,610 (10.7)88,309 (10.4)133,647 (15.9)
       ≥277,295 (4.9)92,885 (7.6)20,029 (6.2)30,529 (3.6)118,286 (14.1)
      Median Income
       First quartile255,028 (16.1)239,943 (19.6)57,660 (17.8)150,495 (17.6)178,925 (21.3)
       Second quartile338,028 (21.4)273,819 (22.4)70,557 (21.8)195,940 (23.0)202,312 (24.1)
       Third quartile414,048 (26.2)308,016 (25.2)82,412 (25.4)224,552 (26.3)219,744 (26.2)
       Fourth quartile538,267 (34.1)373,848 (30.6)104,758 (32.3)261,682 (30.7)220,139 (26.3)
       Missing34,947 (2.2)26,444 (2.2)8727 (2.7)20,501 (2.4)17,146 (2.0)
      Patient location
       Urban1,405,223 (88.9)1,087,846 (89.0)286,143 (88.3)752,715 (88.2)750,536 (89.5)
       Large rural town85,973 (5.4)68,015 (5.6)18,880 (5.8)49,330 (5.8)46,026 (5.5)
       Small rural town29,173 (1.8)23,322 (1.9)5472 (1.7)16,912 (2.0)13,754 (1.6)
       Isolated rural town21,382 (1.4)13,994 (1.1)3979 (1.2)10,921 (1.3)8829 (1.1)
       Missing38,567 (2.4)28,893 (2.4)9640 (3.0)23,292 (2.7)19,121 (2.3)
      Number of chronic conditions
       Average0.8 (1.1)1.0 (1.3)0.9 (1.1)1.5 (1.6)2.3 (2.0)
      Table 2Complications Within 30 Days of Screening/Surveillance Colonoscopies vs Non-Screening/Surveillance Colonoscopies
      Raw rates and adjusted logistic regression for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      CharacteristicColonoscopy for screening/surveillanceColonoscopy for other than screening/surveillance
      Diagnostic
      Adjusted logistic regression between screening/surveillance diagnostic colonoscopy vs non-screening/surveillance diagnostic colonoscopy.
      (n = 645,095)
      Biopsy or intervention
      Adjusted logistic regression between screening/surveillance biopsy/intervention colonoscopy vs non-screening/surveillance biopsy/intervention colonoscopy.
      (n = 935,223)
      Total (n = 1,580,318)Diagnostic (n = 737,048)Biopsy or intervention (n = 485,022)Total (n = 1,222,070)
      Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)
      Any complication7048109 (107–112)
      Adjusted logistic regression with P < .001.
      20,365218 (215–221)
      Adjusted logistic regression with P < .001.
      27,413173 (171–176)17,625239 (236–243)19,413400 (395–406)37,038303 (300–306)
      GI291845 (44–47)
      Adjusted logistic regression with P < .001.
      9768104 (102–107)
      Adjusted logistic regression with P < .001.
      12,68680 (79–82)8841120 (117–122)9547197 (193–201)18,388150 (148–153)
      Serious (GI)129520 (19–21)
      Adjusted logistic regression with P < .001.
      609465 (64–67)
      Adjusted logistic regression with P < .001.
      738946.8 (45.7–47.8)451561 (59–63)5031104 (101–107)954678.1 (76.6–79.7)
       Lower GI bleeding3445.3 (4.8–5.9)
      Adjusted logistic regression with P < .001.
      340136.4 (35.1–37.6)
      Adjusted logistic regression with P < .001.
      374523.7 (22.9–24.5)139018.9 (17.9–19.8)243450.2 (48.2–52.2)382431.3 (30.3–32.3)
       Upper GI bleeding610.9 (0.7–1.2)
      Adjusted logistic regression with P < .001.
      1711.8 (1.6–2.1)
      Adjusted logistic regression with P < .01.
      2321.5 (1.3–1.7)2042.8 (2.4–3.1)1593.3 (2.8–3.8)3633.0 (2.7–3.3)
       Perforation1862.9 (2.5–3.3)
      Adjusted logistic regression with P < .001.
      5866.3 (5.8–6.8)
      Adjusted logistic regression with P < .001.
      7724.9 (4.5–5.2)3534.8 (4.3–5.3)52510.8 (9.9–11.7)8787.2 (6.7–7.7)
       Colonic diverticulitis80812.5 (11.7–13.4)
      Adjusted logistic regression with P < .001.
      289531.0 (29.8–32.1)
      Adjusted logistic regression with P < .001.
      370323.4 (22.7–24.2)302041.0 (39.5–42.4)247551.0 (49.0–53.0)549545.0 (43.8–46.2)
       Small bowel diverticulitis
      Adjusted logistic regression did not converge for small intestine diverticulitis.
      70.1 (0.0–0.2)210.2 (0.1–0.3)280.2 (0.1–0.2)300.4 (0.3–0.6)290.6 (0.4–0.8)590.5 (0.4–0.6)
      Non-serious (GI)
      Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      176927 (26–29)
      Adjusted logistic regression with P < .001.
      422745 (44–47)
      Adjusted logistic regression with P < .001.
      599637.9 (37.0–38.9)491667 (65–69)5195107 (104–110)10,11182.7 (81.1–84.3)
       Abdominal pain119218.5 (17.4–19.5)
      Adjusted logistic regression with P < .001.
      295231.6 (30.4–32.7)
      Adjusted logistic regression with P < .001.
      414426.2 (25.4–27.0)357748.5 (46.9–50.1)350972.3 (70.0–74.7)708658.0 (56.6–59.3)
       Nausea6149.5 (8.8–10.3)
      Adjusted logistic regression with P < .001.
      127513.6 (12.9–14.4)
      Adjusted logistic regression with P < .001.
      188912.0 (11.4–12.5)146919.9 (18.9–20.9)154431.8 (30.2–33.4)301324.7 (23.8–25.5)
       Gastroenteritis (noninfectious)1592.5 (2.1–2.8)
      Adjusted logistic regression with P < .001.
      4725.0 (4.6–5.5)
      Adjusted logistic regression with P < .001.
      6314.0 (3.7–4.3)4906.6 (6.1–7.2)92619.1 (17.9–20.3)141611.6 (11.0–12.2)
      Non-GI491576 (74–78)
      Adjusted logistic regression with P < .001.
      14,316153 (151–156)
      Adjusted logistic regression with P < .001.
      19,231122 (120–123)11,623158 (155–161)13,400276 (272–281)25,023205 (202–207)
      Cardiac291145 (43–47)
      Adjusted logistic regression with P < .001.
      858892 (90–94)
      Adjusted logistic regression with P < .001.
      11,49973 (71–74)667891 (88–93)7244149 (146–153)13,922114 (112–116)
       Acute MI1602.5 (2.1–2.9)
      Adjusted logistic regression with P < .025.
      3974.2 (3.8–4.7)
      Adjusted logistic regression with P < .001.
      5573.5 (3.2–3.8)3124.2 (3.8–4.7)3346.9 (6.1–7.6)6465.3 (4.9–5.7)
       Coronary atherosclerosis154223.9 (22.7–25.1)
      Adjusted logistic regression with P < .001.
      478051.1 (49.7–52.6)
      Adjusted logistic regression with P < .001.
      632240.0 (39.0–41.0)373550.7 (49.1–52.3)390380.5 (78.0–83.0)763862.5 (61.1–63.9)
       Congestive heart failure6249.7 (8.9–10.4)
      Adjusted logistic regression with P < .001.
      192720.6 (19.7–21.5)
      Adjusted logistic regression with P < .001.
      255116.1 (15.5–16.8)177324.1 (22.9–25.2)185938.3 (36.6–40.1)363229.7 (28.8–30.7)
       Cardiac dysrhythmia147322.8 (21.7–24.0)
      Adjusted logistic regression with P < .001.
      427245.7 (44.3–47.0)
      Adjusted logistic regression with P < .001.
      574536.4 (35.4–37.3)3,19143.3 (41.8–44.8)362674.8 (72.3–77.2)681755.8 (54.5–57.1)
       Cardiac structural2994.6 (4.1–5.2)
      Adjusted logistic regression with P < .001.
      9299.9 (9.3–10.6)
      Adjusted logistic regression with P < .001.
      12287.8 (7.3–8.2)84611.5 (10.7–12.3)90718.7 (17.5–19.9)175314.3 (13.7–15.0)
      Cerebrovascular70711.0 (10.2–11.8)
      Adjusted logistic regression with P < .001.
      198921.3 (20.3–22.2)
      Adjusted logistic regression with P < .001.
      269617.1 (16.4–17.7)166322.6 (21.5–23.6)165134.0 (32.4–35.7)331427.1 (26.2–28.0)
       Ischemic stroke3004.7 (4.1–5.2)
      Adjusted logistic regression with P < .001.
      8489.1 (8.5–9.7)
      Adjusted logistic regression with P < .001.
      11487.3 (6.8–7.7)7069.6 (8.9–10.3)73615.2 (14.1–16.3)144211.8 (11.2–12.4)
       Hemorrhagic stroke280.4 (0.3–0.6)
      Adjusted logistic regression with P < .025.
      650.7 (0.5–0.9)
      Adjusted logistic regression with P > .05.
      930.6 (0.5–0.7)731.0 (0.8–1.2)330.7 (0.4–0.9)1060.9 (0.7–1.0)
       TIA4296.7 (6.0–7.3)
      Adjusted logistic regression with P < .001.
      124213.3 (12.5–14.0)
      Adjusted logistic regression with P < .001.
      167110.6 (10.1–11.1)104214.1 (13.3–15.0)101420.9 (19.6–22.2)205616.8 (16.1–17.6)
      Pulmonary118218.3 (17.3–19.4)
      Adjusted logistic regression with P < .001.
      371939.8 (38.5–41.0)
      Adjusted logistic regression with P < .001.
      490131 (30–32)303841.2 (39.8–42.7)328667.7 (65.4–70.1)632452 (50–53)
       Pneumonia4416.8 (6.2–7.5)
      Adjusted logistic regression with P < .001.
      118512.7 (11.9–13.4)
      Adjusted logistic regression with P < .001.
      162610.3 (9.8–10.8)111415.1 (14.2–16.0)115923.9 (22.5–25.3)227318.6 (17.8–19.4)
       Bronchitis87613.6 (12.7–14.5)
      Adjusted logistic regression with P < .001.
      292731.3 (30.2–32.4)
      Adjusted logistic regression with P < .001.
      380324.1 (23.3–24.8)224230.4 (29.2–31.7)245150.5 (48.5–52.5)469338.4 (37.3–39.5)
      Infection
      Includes bacteremia and fever.
      3395.3 (4.7–5.8)
      Adjusted logistic regression with P < .001.
      103711.1 (10.4–11.8)
      Adjusted logistic regression with P < .001.
      13768.7 (8.2–9.2)7309.9 (9.2–10.6)113123.3 (22.0–24.7)186115.2 (14.5–15.9)
       Bacteremia330.5 (0.3–0.7)
      Adjusted logistic regression with P > .05.
      911.0 (0.8–1.2)
      Adjusted logistic regression with P < .01.
      1240.8 (0.6–0.9)821.1 (0.9–1.4)821.7 (1.3–2.1)1641.3 (1.1–1.5)
      Death1332.1 (1.7–2.4)
      Adjusted logistic regression with P < .001.
      3794.1 (3.6–4.5)
      Adjusted logistic regression with P < .001.
      5123.2 (3.0–3.5)3675.0 (4.5–5.5)4429.1 (8.3–10.0)8096.6 (6.2–7.1)
      GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a Raw rates and adjusted logistic regression for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      b Adjusted logistic regression did not converge for small intestine diverticulitis.
      c Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      d Includes bacteremia and fever.
      e Adjusted logistic regression between screening/surveillance diagnostic colonoscopy vs non-screening/surveillance diagnostic colonoscopy.
      f Adjusted logistic regression between screening/surveillance biopsy/intervention colonoscopy vs non-screening/surveillance biopsy/intervention colonoscopy.
      g Adjusted logistic regression with P < .001.
      h Adjusted logistic regression with P < .01.
      i Adjusted logistic regression with P < .025.
      j Adjusted logistic regression with P > .05.

       Demographics: Comparator Groups

      The 3 comparator groups included a total of 2,015,550 patients (Table 1). The majority of patients undergoing joint injection/aspiration/lithotripsy and arthroscopies/carpal tunnel release were under age 60, while the majority of those undergoing cataract operations were over age 70 (Table 1). Patients undergoing cataract operations were nearly twice as likely to have at least 1 comorbidity documented vs those undergoing joint injection/aspiration/lithotripsy or arthroscopy/carpal tunnel release (Table 1).

       30-day Complication Rates: Screening/Surveillance vs Non-Screening/Surveillance Colonoscopy

      Complications were documented within 30 days in 109 (95% CI 107–112) per 10,000 S-Colo-Diag patients, and in 218 (95% CI 215–221) per 10,000 S-Colo-Int patients (Table 2). The rates were approximately double in NS-Colo patients (Table 2).
      The rates of lower GI bleeding were 5.3 (95% CI 4.8–5.9] per 10,000 S-Colo-Diag patients, and 36.4 (95% CI 35.1–37.6) per 10,000 S-Colo-Int patients (Table 2). The rates were substantially higher in NS-Colo patients (Table 2, Figure 1). The rates of perforation were 2.9 (95% CI 2.5–3.3) per 10,000 S-Colo-Diag patients, and 6.3 (95% CI 5.8–6.8) per 10,000 S-Colo-Int patients (Table 2). Again, the rates were substantially higher in NS-Colo patients (Table 2, Figure 1). Hospitalization rates were 50.8%–70.7% among patients who developed lower GI bleeding, and 98.9%–100% among patients who developed perforation (Table 3). In contrast, hospitalization in patients with a diagnosis of abdominal pain or nausea was comparatively rare (Table 3).
      Figure thumbnail gr1
      Figure 1Crude rates (and 95% CIs) of selected severe GI and non-GI complications after colonoscopy, by procedure subtype, and after non-colonoscopy comparators presumed to have low systemic risk. For colonoscopy, “diagnostic” signifies without biopsy/polypectomy/intervention, and “intervention” signifies with biopsy/polypectomy/intervention. Serious GI complication rates were higher after colonoscopy than after comparator procedures, as might be expected. Serious non-GI complications tended not to be higher after screening/surveillance colonoscopy than after comparator procedures. Adjusted odds ratios are shown in .
      Table 3Percent of 30-Day Complications Leading to Hospitalization, and Occurring Within 7, 14, and 21 Days of a Colonoscopy
      CharacteristicColonoscopy for screening/surveillanceColonoscopy for other than screening/surveillance
      DiagnosticBiopsy or interventionDiagnosticBiopsy or intervention
      Percent of 30 day complicationsPercent of 30 day complicationsPercent of 30 day complicationsPercent of 30 day complications
      HospitalizedWithin 7 daysWithin 14 daysWithin 21 daysHospitalizedWithin 7 daysWithin 14 daysWithin 21 daysHospitalizedWithin 7 daysWithin 14 daysWithin 21 daysHospitalizedWithin 7 daysWithin 14 daysWithin 21 days
      Any complication57.635.455.675.765.338.059.577.557.437.958.477.367.034.856.777.0
      GI41.344.161.979.153.950.770.884.041.946.164.780.950.345.465.381.5
      Serious (GI)70.547.565.081.473.053.275.286.864.951.267.882.573.746.866.782.8
       Lower GI bleeding66.035.555.579.170.756.483.692.250.851.969.483.464.451.272.586.6
       Upper GI bleeding96.732.860.777.098.226.354.474.993.131.457.878.494.332.755.378.0
       Perforation100.061.374.787.699.563.174.284.699.459.873.784.498.941.959.678.5
       Colonic diverticulitis65.851.667.881.173.550.370.983.965.053.368.582.776.146.766.481.7
       Small bowel diverticulitis71.442.971.485.785.733.361.971.493.346.756.776.786.227.637.972.4
      Nonserious (GI)
      Includes nausea, abdominal pain and noninfectious gastroenteritis.
      19.641.359.477.324.846.764.279.620.540.560.678.926.345.064.480.5
       Abdominal pain11.045.161.177.514.052.769.182.512.743.162.279.614.147.266.981.9
       Nausea29.234.954.974.437.435.554.773.630.235.757.777.835.835.355.176.2
       Gastroenteritis (noninfectious)39.025.251.676.151.937.357.676.342.232.951.274.549.944.664.880.2
      Non-GI69.730.452.073.876.831.454.674.771.832.754.474.981.528.351.774.3
      Cardiac72.031.052.373.578.631.053.573.975.831.553.974.383.727.250.773.6
       Acute MI96.332.556.377.597.534.053.974.698.132.152.277.698.230.249.472.5
       Coronary atherosclerosis76.328.750.873.581.729.653.373.278.831.353.474.685.526.950.673.5
       Congestive heart failure83.328.450.272.385.128.351.473.684.528.553.173.788.526.349.473.4
       Cardiac dysrhythmia70.732.552.773.778.931.852.473.876.031.553.773.284.827.150.072.7
       Cardiac structural90.027.849.273.292.627.151.172.091.828.352.873.592.725.749.272.9
      Cerebrovascular69.328.949.974.177.130.353.573.374.230.453.374.183.226.249.674.3
       Ischemic stroke83.326.344.370.389.428.151.973.686.030.053.374.290.925.747.772.0
       Hemorrhagic stroke85.721.439.385.789.223.144.667.790.424.750.775.390.921.245.566.7
       TIA59.229.452.274.868.531.154.272.566.429.852.874.177.125.849.574.4
      Pulmonary67.525.547.069.675.926.249.571.772.927.850.272.380.324.948.072.4
       Pneumonia83.023.644.269.287.826.448.771.285.725.249.270.292.823.446.571.0
       Bronchitis62.425.747.869.273.425.849.472.168.928.350.273.276.125.148.072.8
      Infection
      Includes bacteremia and fever.
      69.931.654.976.177.034.652.871.674.728.549.072.381.126.446.871.4
       Bacteremia93.918.242.484.896.726.445.163.795.124.456.181.796.326.846.372.0
      Death91.724.139.869.987.323.547.066.891.325.147.170.092.323.845.570.1
      GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a Includes nausea, abdominal pain and noninfectious gastroenteritis.
      b Includes bacteremia and fever.
      The rates of serious non-GI complications, as well as additional GI complications, are shown in Table 2, and the fractions of these complications leading to hospitalization are shown in Table 3. Selected non-GI complications are highlighted in Figure 1. Non-GI complication rates per 10,000 S-Colo patients included 0.6 (95% CI 0.5–0.7) for hemorrhagic stroke, 0.8 (95% CI 0.6–0.9) for bacteremia, 3.5 (95% CI 3.2–3.8) for MI, 7.3 (95% CI 6.8–7.7) for ischemic stroke, 10.3 (95% CI 9.8–10.8) for pneumonia, and 36.4 (95% CI 35.4–37.3) for cardiac dysrhythmia (Table 2). The rates were higher after S-Colo-Int vs S-Colo-Diag, and after NS-Colo vs S-Colo (Table 2, Figure 1). The vast majority of these serious non-GI complications resulted in hospitalization: 85.7%–90.9% for hemorrhagic stroke, 93.9%–96.7% for bacteremia, 96.3%–98.2% for MI, 83.3%–90.9% for ischemic stroke, 83.0%–92.8% for pneumonia, and 70.7%–84.8% for cardiac dysrhythmia (Table 3).
      After adjusting for age, demographic and social factors, and comorbidities, nearly all GI and non-GI complications were less likely after S-Colo than after NS-Colo (Supplementary Table 6).

       Time-course of Complications: 7-day, 14-day and 21-day Complication Rates

      The fractions of 30-day serious and non-serious GI complications that occurred within 7, 14, and 21 days exceeded the approximately 25%, 50%, and 75%, respectively, that would be expected if the complications rates were constant over 30 days post-colonoscopy (Table 3). For instance, 35.5%–56.4% of lower GI bleeds and 41.9%–63.1% of perforations after colonoscopy presented within 7 days, and 55.5%–83.6% of lower GI bleeds and 59.6%–74.7% of perforations after colonoscopy presented within 14 days (Table 3).
      In contrast, the fractions of 30-day serious non-GI complications that occurred within 7 days tended to exceed 25% slightly, but the fractions that occurred within 14 and 21 days tended to approximate 50% and 75%, respectively (Table 3).

       30-day Complication Rates: Colonoscopy vs Comparator Procedures

      After comparator procedures, complication rates ranged from 150 (95% CI 147–153) to 280 (95% CI 275–286) per 10,000 patients (Supplementary Table 7A). Approximately one half of post-colonoscopy complications were GI vs approximately one fourth to one third after comparator procedures, and serious GI complications were substantially more likely after colonoscopy vs comparator procedures (Table 2, Supplementary Table 7A, Figure 1). The unadjusted rates for non-GI complications after S-Colo tended to be higher than those after arthroscopy/carpal tunnel release, and lower than those after joint injection/aspiration/lithotripsy and cataract surgery (Table 2, Supplementary Table 7A, Figure 1). The lithotripsy subgroup tended to have higher unadjusted rates for GI complications, but not for non-GI complications, than the joint injection and joint aspiration subgroups (Supplementary Table 7B).
      After adjusting for age, demographic and social factors, and comorbidities, the odds of serious GI complications after S-Colo ranged from 2.18 (95% CI 2.02–2.36) compared with joint injection/aspiration/lithotripsy to 5.13 (95% CI 4.81–5.47) compared with cataract operations (Table 4).
      Table 4Odds of Complications After Screening/Surveillance Colonoscopy Compared With Non-colonoscopy Procedures Presumed to Have Low Systemic Risk
      Adjusted for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      ComplicationsOdds ratios (95% CI)
      Joint injections, aspirations, lithotripsyArthroscopies, carpal tunnel surgeryCataract surgery
      Any complication0.64 (0.63–0.66)0.99 (0.96–1.01)1.41 (1.37–1.44)
      GI complication0.84 (0.81–0.88)1.67 (1.61–1.74)2.66 (2.55–2.77)
      Serious GI complication
      Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      2.18 (2.02–2.36)2.85 (2.67–3.04)5.13 (4.81–5.47)
       Lower GI bleeding3.43 (3.00–3.92)2.33 (2.14–2.55)5.84 (5.29–6.45)
       GI perforation2.13 (1.67–2.72)7.87 (5.67–10.93)10.39 (7.93–13.61)
      Non-serious GI complication
      Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      0.48 (0.46–0.51)1.10 (1.04–1.15)1.50 (1.42–1.59)
      Non-GI0.63 (0.61–0.65)0.98 (0.95–1.01)1.26 (1.22–1.29)
      Cardiac0.62 (0.59–0.64)1.17 (1.12–1.22)1.22 (1.18–1.26)
       Acute MI0.67 (0.56–0.81)0.99 (0.83–1.19)0.85 (0.74–0.97)
       CHF0.70 (0.64–0.76)1.31 (1.19–1.44)0.96 (0.90–1.02)
       Cardiac dysrhythmia0.69 (0.65–0.73)1.21 (1.14–1.28)1.33 (1.27–1.40)
      Cerebrovascular0.61 (0.56–0.66)1.08 (0.99–1.18)1.05 (0.99–1.12)
       Ischemic stroke0.66 (0.59–0.75)1.13 (0.99–1.29)1.00 (0.91–1.10)
       Hemorrhagic stroke0.70 (0.44–1.11)0.94 (0.60–1.47)0.76 (0.55–1.04)
       TIA0.59 (0.53–0.65)1.03 (0.93–1.15)1.08 (1.00–1.17)
      Pulmonary0.64 (0.61–0.68)1.05 (0.98–1.11)1.03 (0.98–1.08)
       Pneumonia0.65 (0.58–0.72)0.85 (0.77–0.94)0.95 (0.87–1.03)
       Bronchitis0.64 (0.60–0.69)1.18 (1.10–1.27)1.04 (0.98–1.10)
      Infection
      Includes bacteremia and fever.
      0.85 (0.75–0.96)0.36 (0.33–0.39)1.80 (1.60–2.02)
       Bacteremia0.38 (0.28–0.53)0.87 (0.61–1.24)1.03 (0.75–1.40)
      Death0.77 (0.63–0.94)1.64 (1.29–2.08)1.09 (0.94–1.27)
      CHF, congestive heart failure; GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a Adjusted for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      b Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      c Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      d Includes bacteremia and fever.
      In contrast to these higher adjusted odds of serious GI complication, serious non-GI complications were generally less likely after S-Colo compared with joint injection/aspiration/lithotripsy (when combined into a single comparator group [Table 4], or when separated into 3 subgroups [Supplementary Table 8]), and were generally comparable or slightly more likely compared with arthroscopy/carpal tunnel release or cataract operations (Table 4). For instance, the adjusted odds of MI after S-Colo were 0.67 (95% CI 0.56–0.81) compared with joint injection/aspiration/lithotripsy, and 0.99 (95% CI 0.83–1.19) compared with arthroscopy/carpal tunnel release (Table 4). Similarly, the adjusted odds of ischemic stroke and serious pulmonary events after S-Colo vs comparator procedures ranged, respectively, from 0.66 (95% CI 0.59–0.75) to 1.13 (95% CI 0.99–1.29), and from 0.64 (95% CI 0.61–0.68) to 1.05 (95% CI 0.98–1.11) (Table 4).
      In contrast to these findings for S-Colo, non-GI complications after NS-Colo tended to be lower or comparable to rates after joint injection/aspiration/lithotripsy, and were generally higher compared with rates after arthroscopy/carpal tunnel release or cataract operations (Supplementary Table 9).
      Similar results were seen after eliminating colonoscopies that were within ±30 days of EGD, sigmoidoscopy, or small bowel endoscopy.

       Risk Factors for 30-day Complications After Colonoscopy

      The multivariable, adjusted predictors of post-colonoscopy complications are shown in Table 5. Compared with S-Colo, NS-Colo was associated with higher odds of complications, ranging from 1.26 (95% CI 1.19–1.33) for cerebrovascular complications to 1.58 (95% CI 1.52–1.63) for serious GI complications (Table 5). Biopsy/intervention was associated with higher odds of serious GI as well as non-GI complications (Table 5).
      Table 5Risk Factors for Post-Colonoscopy Complications Odds Ratios (95% CI)
      ComplicationsAny complicationGI complicationsSerious GI complicationsLower GI bleedingGI perforationNon-serious

      GI complications
      CardiacCerebro vascularPulmonary
      Age
       <501.60 (1.56–1.65)2.04 (1.97–2.11)1.38 (1.31–1.46)1.55 (1.44–1.68)1.19 (0.99–1.43)2.52 (2.42–2.63)0.75 (0.70–0.79)0.62 (0.54–0.72)0.96 (0.89–1.05)
       50–59---------
       60–691.18 (1.15–1.21)1.02 (0.99–1.06)1.25 (1.19–1.31)1.17 (1.09–1.26)1.43 (1.23–1.67)0.82 (0.78–0.87)1.64 (1.57–1.71)1.80 (1.64–1.98)1.30 (1.22–1.38)
       70–791.57 (1.53–1.62)1.27 (1.21–1.32)1.76 (1.66–1.86)1.56 (1.44–1.70)2.39 (2.01–2.85)0.85 (0.80–0.90)2.61 (2.49–2.74)2.88 (2.61–3.18)1.85 (1.73–1.99)
       ≥802.43 (2.35–2.51)1.85 (1.76–1.94)2.85 (2.68–3.03)2.70 (2.47–2.96)2.83 (2.32–3.45)0.98 (0.91–1.06)4.68 (4.45–4.92)5.46 (4.92–6.05)2.81 (2.61–3.02)
      Gender
       Male---------
       Female0.98 (0.96–1.00)1.11 (1.08–1.13)0.84 (0.82–0.87)0.71 (0.68–0.75)1.16 (1.05–1.28)1.51 (1.46–1.56)0.74 (0.72–0.76)0.90 (0.86–0.95)0.89 (0.86–0.92)
      Race
       White---------
       Black1.26 (1.21–1.30)1.20 (1.14–1.26)1.18 (1.09–1.26)1.42 (1.28–1.57)0.95 (0.74–1.22)1.21 (1.13–1.30)1.18 (1.11–1.24)1.82 (1.65–2.01)1.09 (1.01–1.18)
       Hispanic0.97 (0.94–0.99)1.09 (1.06–1.12)1.10 (1.05–1.15)1.12 (1.05–1.20)0.82 (0.70–0.95)1.08 (1.03–1.12)0.78 (0.75–0.81)0.97 (0.90–1.04)0.64 (0.60–0.68)
       Asian/Pacific Islander0.75 (0.73–0.78)0.76 (0.73–0.80)0.85 (0.80–0.90)1.17 (1.08–1.26)0.85 (0.71–1.02)0.67 (0.63–0.71)0.69 (0.66–0.73)0.79 (0.72–0.87)0.54 (0.50–0.59)
       Native American1.41 (1.17–1.70)1.50 (1.17–1.93)1.17 (0.78–1.76)1.46 (0.84–2.52)0.48 (0.07–3.44)1.83 (1.37–2.44)1.19 (0.84–1.67)1.70 (0.93–3.09)1.44 (0.97–2.14)
       Other0.71 (0.67–0.76)0.77 (0.71–0.84)0.78 (0.69–0.88)0.94 (0.80–1.10)0.60 (0.39–0.90)0.76 (0.68–0.85)0.70 (0.63–0.78)0.70 (0.57–0.87)0.51 (0.43–0.60)
      Charlson-Deyo score
       0---------
       12.06 (2.01–2.12)1.75 (1.69–1.81)1.57 (1.50–1.65)1.60 (1.49–1.72)1.79 (1.53–2.11)1.97 (1.88–2.07)2.38 (2.29–2.47)2.49 (2.31–2.69)3.38 (3.19–3.57)
       ≥25.31 (5.20–5.42)3.31 (3.20–3.42)3.21 (3.08–3.35)3.70 (3.48–3.93)5.39 (4.76–6.11)3.22 (3.08–3.38)7.05 (6.83–7.27)6.92 (6.50–7.37)8.75 (8.34–9.17)
      Median income
       First quartile1.13 (1.11–1.16)1.10 (1.06–1.14)1.08 (1.03–1.14)1.11 (1.03–1.19)0.88 (0.75–1.03)1.11 (1.06–1.17)1.14 (1.10–1.19)1.12 (1.03–1.21)1.58 (1.49–1.68)
       Second quartile1.15 (1.12–1.17)1.14 (1.10–1.18)1.13 (1.08–1.18)1.15 (1.07–1.23)1.05 (0.92–1.21)1.15 (1.10–1.21)1.15 (1.11–1.19)1.16 (1.07–1.25)1.49 (1.41–1.58)
       Third quartile1.09 (1.07–1.12)1.08 (1.05–1.12)1.08 (1.03–1.12)1.07 (1.01–1.14)1.01 (0.89–1.16)1.10 (1.05–1.15)1.08 (1.04–1.12)1.09 (1.02–1.18)1.23 (1.16–1.30)
       Fourth quartile---------
      Insurance
       Medicare1.45 (1.42–1.49)1.30 (1.26–1.34)1.14 (1.09–1.20)1.16 (1.08–1.24)0.99 (0.86–1.14)1.51 (1.44–1.59)1.60 (1.54–1.66)1.90 (1.75–2.06)2.09 (1.97–2.22)
       Medicaid1.92 (1.86–1.98)1.92 (1.85–2.00)1.47 (1.38–1.57)1.52 (1.39–1.67)0.93 (0.72–1.19)2.29 (2.18–2.41)1.78 (1.67–1.89)2.41 (2.14–2.71)3.00 (2.78–3.24)
       Private---------
       Self-pay1.53 (1.43–1.63)1.47 (1.35–1.61)1.41 (1.25–1.60)1.32 (1.10–1.59)1.43 (0.98–2.07)1.50 (1.33–1.69)1.53 (1.37–1.71)1.59 (1.24–2.03)1.73 (1.46–2.06)
       Other1.35 (1.30–1.42)1.37 (1.30–1.46)1.19 (1.09–1.30)1.39 (1.24–1.57)1.01 (0.76–1.35)1.60 (1.48–1.72)1.25 (1.15–1.36)1.47 (1.23–1.75)1.68 (1.49–1.88)
      Biopsy or intervention1.52 (1.49–1.54)1.61 (1.57–1.65)1.82 (1.76–1.88)3.07 (2.91–3.25)1.89 (1.70–2.11)1.43 (1.38–1.47)1.39 (1.35–1.43)1.28 (1.21–1.35)1.41 (1.35–1.46)
      Other endoscopy within 30 days1.33 (1.31–1.36)1.25 (1.22–1.29)0.99 (0.95–1.03)1.16 (1.10–1.23)0.81 (0.71–0.92)1.56 (1.50–1.62)1.31 (1.27–1.35)1.29 (1.21–1.37)1.26 (1.21–1.32)
      Indication for colonoscopy
       Screening/surveillance---------
       Non-screening/surveillance1.46 (1.44–1.49)1.52 (1.48–1.56)1.58 (1.52–1.63)1.31 (1.24–1.37)1.52 (1.37–1.69)1.51 (1.46–1.57)1.34 (1.30–1.37)1.26 (1.19–1.33)1.36 (1.30–1.41)
      Patient location
       Urban---------
       Large rural town1.09 (1.06–1.13)1.22 (1.16–1.27)1.15 (1.09–1.23)1.20 (1.10–1.32)1.36 (1.13–1.65)1.27 (1.20–1.35)0.95 (0.91–1.01)0.94 (0.85–1.05)1.08 (1.00–1.16)
       Small rural town1.07 (1.01–1.13)1.21 (1.12–1.30)1.14 (1.03–1.26)1.15 (0.99–1.34)1.32 (0.96–1.81)1.30 (1.17–1.43)0.87 (0.80–0.95)1.07 (0.90–1.26)1.03 (0.91–1.15)
       Isolated rural town1.03 (0.96–1.10)1.17 (1.07–1.29)1.14 (1.00–1.29)1.11 (0.91–1.35)1.37 (0.94–2.01)1.20 (1.06–1.37)0.83 (0.74–0.93)0.80 (0.63–1.02)1.02 (0.88–1.18)
      Number of chronic conditions1.02 (1.02–1.03)1.07 (1.06–1.08)1.14 (1.13–1.15)1.07 (1.06–1.09)0.96 (0.93–1.00)0.99 (0.98–1.00)1.03 (1.02–1.04)1.03 (1.01–1.04)1.00 (0.99–1.01)
      C statistic0.7470.7160.7250.7330.6050.7300.8200.8120.821
      Both younger and older patients had higher odds of post-colonoscopy complications compared with persons ages 50–59, except for pulmonary, cardiac, and cerebrovascular complications, which increased progressively with age (Table 5 and Figure 2). Compared with men, women were more likely to experience perforation, but less likely to experience lower GI bleeding or serious cardiac, cerebrovascular, or pulmonary events (Table 5). Compared with white patients, black patients had higher odds, and Asian patients had lower odds for most complications; Hispanic patients had higher odds for some and lower odds for other complications (Table 5). Comorbidities increased the odds of complications, with 2 or more comorbidities increasing the odds for all complications substantially (maximum 8.75 [95% CI 8.34–9.17] for pulmonary events). Lower incomes and government insurance or self-pay status vs private insurance were associated with higher odds of complications (Table 5).
      Figure thumbnail gr2
      Figure 2Post-colonoscopy complications by age. Patient age groups younger and older than the 50–59 years reference group had higher odds of post-colonoscopy complications, in particular GI complications, except for cardiac and cerebrovascular complications, which increased progressively with age. The fact that most colonoscopies in those under age 50 were likely prompted by lower GI symptoms and possible underlying serious GI disease, such as inflammatory bowel disease, may explain the higher complication rates in this younger group.

      Discussion

      Our large, population-based study of post-colonoscopy complications in California’s diverse population addresses key knowledge gaps regarding the risks of colonoscopy, particularly the risk of serious non-GI complications, as identified in the recent systematic review for the USPSTF. We examined in detail GI as well as a broad spectrum of non-GI complications after S-Colo vs NS-Colo, with or without biopsy/intervention. Furthermore, the comparisons of event rates and their time-course after colonoscopy vs those after other common outpatient procedures provide insight into the fraction of post-colonoscopy adverse events that may be attributable to colonoscopy itself.
      The lower GI bleeding rates observed in our study of 5.3 (95% CI 4.8–5.9) per 10,000 S-Colo-Diag patients and 36.4 (95% CI 35.1–37.6) per 10,000 S-Colo-Int patients were lower and higher, respectively, than the pooled estimate after screening colonoscopy from the recent systematic review for the USPSTF (8 [95% CI, 5–14] per 10,000).
      • Lin J.
      • Piper M.
      • Perdue L.
      • et al.
      Screening for colorectal cancer: an updated systematic review for the USPSTF.
      The systematic review pooled estimate may reflect a lower rate of biopsy/intervention than we observed in our study. The perforation rates observed in our study of 2.9 (95% CI 2.5–3.3) per 10,000 S-Colo-Diag patients and 6.3 (95% CI 5.8–6.8) per 10,000 S-Colo-Int patients are consistent with the pooled estimate in the recent systematic review of 4 (95% CI, 2–5) per 10,000.
      • Lin J.
      • Piper M.
      • Perdue L.
      • et al.
      Screening for colorectal cancer: an updated systematic review for the USPSTF.
      A majority of lower GI bleeds and nearly all perforations resulted in hospitalization.
      We found that complication rates after NS-Colo were substantially higher than after S-Colo. A recent meta-analysis that included screening, symptom-driven, and therapeutic colonoscopies also showed higher bleeding and perforation rates than in the systematic review for the USPSTF on screening colonoscopies.
      • Reumkens A.
      • Rondagh E.J.A.
      • Bakker C.M.
      • et al.
      Post-colonoscopy complications: a systematic review, time trends, and meta-analysis of population-based studies.
      In our study, biopsy/intervention was an independent predictor of lower GI bleeding and perforation (adjusted odds ratios 3.07 [95% CI 2.91–3.25] and 1.89 [95% CI 1.70–2.11], respectively). Some
      • Reumkens A.
      • Rondagh E.J.A.
      • Bakker C.M.
      • et al.
      Post-colonoscopy complications: a systematic review, time trends, and meta-analysis of population-based studies.
      • Ko C.W.
      • Riffle S.
      • Michaels L.
      • et al.
      Serious complications within 30 days of screening and surveillance colonoscopy are uncommon.
      • Warren J.L.
      Adverse events after outpatient colonoscopy in the Medicare population.
      • Rabeneck L.
      • Paszat L.F.
      • Hilsden R.J.
      • et al.
      Bleeding and perforation after outpatient colonoscopy and their risk factors in usual clinical practice.
      • Levin T.R.
      • Zhao W.
      • Conell C.
      • et al.
      Complications of colonoscopy in an integrated health care delivery system.
      • Basson M.D.
      • Etter L.
      • Panzini L.A.
      Rates of colonoscopic perforation in current practice.
      • Kang H.Y.
      • Kang H.W.
      • Kim S.G.
      • et al.
      Incidence and management of colonoscopic perforations in Korea.
      but not all previous studies
      • Arora G.
      • Mannalithara A.
      • Singh G.
      • et al.
      Risk of perforation from a colonoscopy in adults: a large population-based study.
      have found that adjusted risk increased with polypectomy.
      • Lin J.
      • Piper M.
      • Perdue L.
      • et al.
      Screening for colorectal cancer: an updated systematic review for the USPSTF.
      The 30-day post-procedure observation period that is commonly used to study complications is somewhat arbitrary and is borrowed from the surgical literature. Our analyses of the time-course of complications demonstrate that the majority of serious as well as non-serious GI adverse events after colonoscopy occurred within 14 days. We cannot determine with confidence from the current analyses whether the GI events that occurred 15–30 days post-colonoscopy represent a background rate in the population independent of recent colonoscopy exposure, but the skew in event rates toward the early post-colonoscopy period argues for a causal association between colonoscopy and any early events that occur in excess of a background rate. In contrast, the relatively steady rates of serious non-GI adverse events over the 30 days post-colonoscopy raise the question of whether most of these events reflect the background rates in the population, which are unrelated to colonoscopy, or whether colonoscopy confers a relatively constant level of risk over at least 30 days. These questions merit further research.
      The recent systematic review for the USPSTF found that serious non-GI harms following colonoscopy were not consistently defined or reported in the available literature, and that the vast majority of studies had no comparator arm, making it impossible to determine the fraction of GI and non-GI complications attributable to colonoscopy itself.
      • Lin J.
      • Piper M.
      • Perdue L.
      • et al.
      Screening for colorectal cancer: an updated systematic review for the USPSTF.
      • Rabeneck L.
      • Paszat L.F.
      • Hilsden R.J.
      • et al.
      Bleeding and perforation after outpatient colonoscopy and their risk factors in usual clinical practice.
      • Levin T.R.
      • Zhao W.
      • Conell C.
      • et al.
      Complications of colonoscopy in an integrated health care delivery system.
      • Basson M.D.
      • Etter L.
      • Panzini L.A.
      Rates of colonoscopic perforation in current practice.
      • Polter D.E.
      Risk of colon perforation during colonoscopy at Baylor University Medical Center.
      • Lippert E.
      • Herfarth H.H.
      • Grunert N.
      • et al.
      Gastrointestinal endoscopy in patients aged 75 years and older: risks, complications, and findings—a retrospective study.
      • Kahi C.J.
      • Azzouz F.
      • Juliar B.E.
      • et al.
      Survival of elderly persons undergoing colonoscopy: implications for colorectal cancer screening and surveillance.
      • Rathgaber S.W.
      • Wick T.M.
      Colonoscopy completion and complication rates in a community gastroenterology practice.
      • Zubarik R.
      • Fleischer D.
      • Mastropietro C.
      • et al.
      Prospective analysis of complications 30 days after outpatient colonoscopy.
      • Lüning T.H.
      • Keemers-Gels M.E.
      • Barendregt W.B.
      • et al.
      Colonoscopic perforations: a review of 30,366 patients.
      • Eckardt V.F.
      • Kanzler G.
      • Schmitt T.
      • et al.
      Complications and adverse effects of colonoscopy with selective sedation.
      • Sieg A.
      • Hachmoeller-Eisenbach U.
      • Eisenbach T.
      Prospective evaluation of complications in outpatient GI endoscopy: a survey among German gastroenterologists.
      • Nelson D.B.
      • McQuaid K.R.
      • Bond J.H.
      Procedural success and complications of large-scale screening colonoscopy.
      • Viiala C.H.
      • Zimmerman M.
      • Cullen D.J.E.
      • et al.
      Complication rates of colonoscopy in an Australian teaching hospital environment.
      • Geenen J.E.
      • Schmitt M.G.
      • Wu W.C.
      • et al.
      Major complications of coloscopy: bleeding and perforation.
      • Farley D.R.
      • Bannon M.P.
      • Zietlow S.P.
      • et al.
      Management of colonoscopic perforations.
      • Anderson M.L.
      • Pasha T.M.
      • Leighton J.A.
      Endoscopic perforation of the colon: lessons from a 10-year study.
      • Chukmaitov A.
      • Bradley C.J.
      • Dahman B.
      • et al.
      Association of polypectomy techniques, endoscopist volume, and facility type with colonoscopy complications.
      • Bielawska B.
      • Day A.G.
      • Lieberman D.A.
      • et al.
      Risk factors for early colonoscopic perforation include non-gastroenterologist endoscopists: a multivariable analysis.
      • Singh H.
      • Penfold R.B.
      • DeCoster C.
      • et al.
      Colonoscopy and its complications across a Canadian regional health authority.
      • Adeyemo A.
      • Bannazadeh M.
      • Riggs T.
      • et al.
      Does sedation type affect colonoscopy perforation rates?.
      • Blotière P.-O.
      • Weill A.
      • Ricordeau P.
      • et al.
      Perforations and haemorrhages after colonoscopy in 2010: a study based on comprehensive French health insurance data (SNIIRAM).
      • Cooper G.S.
      • Kou T.D.
      • Rex D.K.
      Complications following colonoscopy with anesthesia assistance: a population-based analysis.
      • Dominitz J.A.
      • Baldwin L.M.
      • Green P.
      • et al.
      Regional variation in anesthesia assistance during outpatient colonoscopy is not associated with differences in polyp detection or complication rates.
      • Hamdani U.
      Risk factors for colonoscopic perforation: a population-based study of 80118 cases.
      Two informative studies included comparator arms. In a study of Medicare beneficiaries, the adjusted risk for a cardiovascular event was significantly greater for the polypectomy group, but not the diagnostic colonoscopy group, as compared with the colonoscopy-free comparator group; individual outcomes such as MI, arrhythmia, or stroke were not reported.
      • Warren J.L.
      Adverse events after outpatient colonoscopy in the Medicare population.
      In a study of a statutory health insurance fund in Germany, the incidence of MI and stroke was similar in the colonoscopy and comparator groups.
      • Stock C.
      • Ihle P.
      • Sieg A.
      • et al.
      Adverse events requiring hospitalization within 30 days after outpatient screening and nonscreening colonoscopies.
      Our detailed analyses address these gaps in the literature. In our analyses that included 3 comparator arms, serious GI complications were more likely after colonoscopy than after non-GI comparator procedures, as would be expected.
      • Lin J.
      • Piper M.
      • Perdue L.
      • et al.
      Screening for colorectal cancer: an updated systematic review for the USPSTF.
      • Arora G.
      • Mannalithara A.
      • Singh G.
      • et al.
      Risk of perforation from a colonoscopy in adults: a large population-based study.
      • Reumkens A.
      • Rondagh E.J.A.
      • Bakker C.M.
      • et al.
      Post-colonoscopy complications: a systematic review, time trends, and meta-analysis of population-based studies.
      • Warren J.L.
      Adverse events after outpatient colonoscopy in the Medicare population.
      • Stock C.
      • Ihle P.
      • Sieg A.
      • et al.
      Adverse events requiring hospitalization within 30 days after outpatient screening and nonscreening colonoscopies.
      • Quintero E.
      • Castells A.
      • Bujanda L.
      • et al.
      Colonoscopy versus fecal immunochemical testing in colorectal-cancer screening.
      More importantly, we found that serious cardiac, cerebrovascular, and pulmonary complications after S-Colo tended not to be higher than after comparator procedures presumed to be at low risk for systemic complications. The higher risks of serious non-GI complications after joint injection/aspiration/lithotripsy than after S-Colo, with the highest risks after lithotripsy, may be attributable to unmeasured comorbidities in patients undergoing these comparator procedures, or to truly higher procedure-related risk, or both. As with serious GI complications, the vast majority of serious non-GI complications resulted in hospitalization.
      The event rates after these comparator procedures may or may not approximate the expected background event rates in the general population. Previously published data on background cardiovascular event rates allow for informative comparisons. In the general US population, the approximate MI incidence rates per 10,000 persons per month for men and women, respectively, are 6.6 and 2.5 for ages 55–64, 9.5 and 4.4 for ages 65–74, and 18.1 and 9.0 for ages 75–84.
      • Mozaffarian D.
      • Benjamin E.J.
      • Go A.S.
      • et al.
      Heart disease and stroke statistics–2015 update: a report from the American Heart Association.
      The respective approximate rates for stroke are 3.1 and 2.1 for ages 55–64, 5.8 and 4.0 for ages 65–74, and 13.6 and 11.6 for ages 75–84.
      • Mozaffarian D.
      • Benjamin E.J.
      • Go A.S.
      • et al.
      Heart disease and stroke statistics–2015 update: a report from the American Heart Association.
      This suggests that the rates of MI observed in our study within 30 days after colonoscopy may be comparable to, but the rates of stroke may be higher than, the expected background rates in the population. One might speculate whether the apparently higher risk of stroke may be related to stopping aspirin or other medications, inherent risk attributed to the procedure, induction of arrhythmias predisposing to stroke, or relative dehydration related to the preparation.
      Multiple previous studies have found increasing risks of serious post-colonoscopy adverse events with increasing age,
      • Lin J.
      • Piper M.
      • Perdue L.
      • et al.
      Screening for colorectal cancer: an updated systematic review for the USPSTF.
      • Ko C.W.
      • Dominitz J.A.
      Complications of colonoscopy: magnitude and management.
      • Warren J.L.
      Adverse events after outpatient colonoscopy in the Medicare population.
      • Chukmaitov A.
      • Bradley C.J.
      • Dahman B.
      • et al.
      Association of polypectomy techniques, endoscopist volume, and facility type with colonoscopy complications.
      • Bielawska B.
      • Day A.G.
      • Lieberman D.A.
      • et al.
      Risk factors for early colonoscopic perforation include non-gastroenterologist endoscopists: a multivariable analysis.
      • Blotière P.-O.
      • Weill A.
      • Ricordeau P.
      • et al.
      Perforations and haemorrhages after colonoscopy in 2010: a study based on comprehensive French health insurance data (SNIIRAM).
      • Cooper G.S.
      • Kou T.D.
      • Rex D.K.
      Complications following colonoscopy with anesthesia assistance: a population-based analysis.
      • Rutter C.M.
      • Johnson E.
      • Miglioretti D.L.
      • et al.
      Adverse events after screening and follow-up colonoscopy.
      • Ranasinghe I.
      • Parzynski C.S.
      • Searfoss R.
      • et al.
      Differences in colonoscopy quality among facilities: development of a post-colonoscopy risk-standardized rate of unplanned hospital visits.
      but most previous studies have not included persons younger than 50. We found that the risks of serious post-colonoscopy GI, cardiac, cerebrovascular, and pulmonary complications all increased with age after age 50, but that serious post-colonoscopy GI events were also more likely in persons younger than 50 compared with those age 50–59 (Table 4 and Figure 2). This J-shaped association has been reported previously in some
      • Hamdani U.
      Risk factors for colonoscopic perforation: a population-based study of 80118 cases.
      • Rutter C.M.
      • Johnson E.
      • Miglioretti D.L.
      • et al.
      Adverse events after screening and follow-up colonoscopy.
      but not in other studies.
      • Ko C.W.
      • Dominitz J.A.
      Complications of colonoscopy: magnitude and management.
      • Arora G.
      • Mannalithara A.
      • Singh G.
      • et al.
      Risk of perforation from a colonoscopy in adults: a large population-based study.
      • Levin T.R.
      • Zhao W.
      • Conell C.
      • et al.
      Complications of colonoscopy in an integrated health care delivery system.
      • Blotière P.-O.
      • Weill A.
      • Ricordeau P.
      • et al.
      Perforations and haemorrhages after colonoscopy in 2010: a study based on comprehensive French health insurance data (SNIIRAM).
      We hypothesize that our observation is explained by the fact that many colonoscopies in those age 50 and older were performed for colorectal neoplasia screening and surveillance,
      • Ladabaum U.
      • Levin Z.
      • Mannalithara A.
      • et al.
      Colorectal testing utilization and payments in a large cohort of commercially insured US adults.
      whereas most colonoscopies in those under age 50 were likely prompted by lower GI symptoms and possible underlying serious GI disease, such as inflammatory bowel disease. As we found in our study, a previous meta-analysis reported that, compared with surveillance/screening colonoscopies, colonoscopies performed to investigate symptoms had higher odds of perforation and bleeding.
      • Reumkens A.
      • Rondagh E.J.A.
      • Bakker C.M.
      • et al.
      Post-colonoscopy complications: a systematic review, time trends, and meta-analysis of population-based studies.
      In our analysis of predictors of post-colonoscopy complications, comorbidity was more strongly associated with complications than was age, as has been reported previously.
      • Warren J.L.
      Adverse events after outpatient colonoscopy in the Medicare population.
      The predictive models were reasonably good (C-statistic 0.716–0.821), with the exception of the analysis for perforation (C-statistic 0.605). We acknowledge that there may have been incomplete ascertainment of comorbidities in our study. However, there is face validity to the observation that the older patients who underwent cataract procedures had substantially more documented comorbidities than the younger patients who underwent colonoscopy. While we may have underestimated absolute comorbidity rates, there is no reason to expect a differential effect among our study groups. Perforation risk may be particularly operator-dependent, a variable that was not accounted for in our study.
      A wide spectrum of post-colonoscopy complications was associated with income, insurance status, and race. These complex associations may reflect the socioeconomic context in which colonoscopy is delivered, as well as the effects of socioeconomic forces on health. It is troubling that black patients had higher adjusted odds of adverse events in nearly all categories, even after adjustment for comorbidities and all other potential predictors available in the dataset. While we accounted for biopsy or polypectomy, we could not adjust for procedure difficulty. In our study, blacks and Hispanics had higher rates of having low income and public insurance. Cost and poor access to insurance are barriers to colonoscopy screening,
      • Holt C.L.
      • Shipp M.
      • Eloubeidi M.
      • et al.
      Use of focus group data to develop recommendations for demographically segmented colorectal cancer educational strategies.
      • Wilkins T.
      • Gillies R.A.
      • Harbuck S.
      • et al.
      Racial disparities and barriers to colorectal cancer screening in rural areas.
      • Bromley E.G.
      • May F.P.
      • Federer L.
      • et al.
      Explaining persistent under-use of colonoscopic cancer screening in African Americans: a systematic review.
      • James A.S.
      • Daley C.M.
      • Greiner K.A.
      Knowledge and attitudes about colon cancer screening among African Americans.
      and it is conceivable that blacks and Hispanics in our study presented with more advanced pathology that increased the risk of adverse events. Our findings highlight the public health challenge that CRC incidence and mortality rates are highest among blacks,
      • Siegel R.
      • DeSantis C.
      • Jemal A.
      Colorectal cancer statistics, 2014.
      but screening rates are the lowest among blacks,
      • Almario C.V.
      • May F.P.
      • Ponce N.A.
      • et al.
      Racial and ethnic disparities in colonoscopic examination of individuals with a family history of colorectal cancer.
      Hispanics,
      • Laiyemo A.O.
      • Thompson N.
      • Williams C.D.
      • et al.
      Race and colorectal cancer screening compliance among persons with a family history of cancer.
      and low-income patients,
      • Decker K.M.
      • Demers A.A.
      • Nugent Z.
      • et al.
      Longitudinal rates of colon cancer screening use in Winnipeg, Canada: the experience of a universal health-care system with an organized colon screening program.
      even when disparities in insurance coverage are limited or do not exist.
      • O'Malley A.S.
      • Forrest C.B.
      • Feng S.
      • et al.
      Disparities despite coverage: gaps in colorectal cancer screening among Medicare beneficiaries.
      • May F.P.
      • Bromley E.G.
      • Reid M.W.
      • et al.
      Low uptake of colorectal cancer screening among African Americans in an integrated Veterans Affairs health care network.
      • Haddad J.D.
      • You D.M.
      Colorectal cancer screening and race in an equal access medical system.
      Our study’s strengths include its large sample size, which approximates a census of all ambulatory colonoscopies performed during the study dates in California, a state with diversity in its population and its facilities performing colonoscopy. Prior studies tended to be limited to specific insurance programs, single institutions or health plans, and specific age groups,
      • Arora G.
      • Mannalithara A.
      • Singh G.
      • et al.
      Risk of perforation from a colonoscopy in adults: a large population-based study.
      • Warren J.L.
      Adverse events after outpatient colonoscopy in the Medicare population.
      • Levin T.R.
      • Zhao W.
      • Conell C.
      • et al.
      Complications of colonoscopy in an integrated health care delivery system.
      • Basson M.D.
      • Etter L.
      • Panzini L.A.
      Rates of colonoscopic perforation in current practice.
      • Polter D.E.
      Risk of colon perforation during colonoscopy at Baylor University Medical Center.
      • Lippert E.
      • Herfarth H.H.
      • Grunert N.
      • et al.
      Gastrointestinal endoscopy in patients aged 75 years and older: risks, complications, and findings—a retrospective study.
      • Kahi C.J.
      • Azzouz F.
      • Juliar B.E.
      • et al.
      Survival of elderly persons undergoing colonoscopy: implications for colorectal cancer screening and surveillance.
      • Rathgaber S.W.
      • Wick T.M.
      Colonoscopy completion and complication rates in a community gastroenterology practice.
      • Zubarik R.
      • Fleischer D.
      • Mastropietro C.
      • et al.
      Prospective analysis of complications 30 days after outpatient colonoscopy.
      • Eckardt V.F.
      • Kanzler G.
      • Schmitt T.
      • et al.
      Complications and adverse effects of colonoscopy with selective sedation.
      • Farley D.R.
      • Bannon M.P.
      • Zietlow S.P.
      • et al.
      Management of colonoscopic perforations.
      • Anderson M.L.
      • Pasha T.M.
      • Leighton J.A.
      Endoscopic perforation of the colon: lessons from a 10-year study.
      • Adeyemo A.
      • Bannazadeh M.
      • Riggs T.
      • et al.
      Does sedation type affect colonoscopy perforation rates?.
      • Cooper G.S.
      • Kou T.D.
      • Rex D.K.
      Complications following colonoscopy with anesthesia assistance: a population-based analysis.
      • Dominitz J.A.
      • Baldwin L.M.
      • Green P.
      • et al.
      Regional variation in anesthesia assistance during outpatient colonoscopy is not associated with differences in polyp detection or complication rates.
      • Hamdani U.
      Risk factors for colonoscopic perforation: a population-based study of 80118 cases.
      • Ranasinghe I.
      • Parzynski C.S.
      • Searfoss R.
      • et al.
      Differences in colonoscopy quality among facilities: development of a post-colonoscopy risk-standardized rate of unplanned hospital visits.
      • Gatto N.M.
      • Frucht H.
      • Sundararajan V.
      • et al.
      Risk of perforation after colonoscopy and sigmoidoscopy: a population-based study.
      • Zafar H.M.
      • Harhay M.O.
      • Yang J.
      • et al.
      Adverse events following computed tomographic colonography compared to optical colonoscopy in the elderly.
      • Hsieh T.K.
      • Hung L.
      • Kang F.C.
      • et al.
      Anesthesia does not increase the rate of bowel perforation during colonoscopy: a retrospective study.
      • Kamath A.S.
      • Iqbal C.W.
      • Sarr M.G.
      • et al.
      Colonoscopic splenic injuries: incidence and management.
      • Tam M.S.
      • Abbas M.A.
      Perforation following colorectal endoscopy: what happens beyond the endoscopy suite?.
      limiting the generalizability of the results. We examined a wide spectrum of individual complications and included 3 comparator groups. Previous studies tended to focus on serious GI complications only, and most studies lacked comparator arms. Previous studies with sample sizes similar to ours had a primary focus other than adverse events or focused only on specific complications, and did not include comparator groups.
      • Chukmaitov A.
      • Bradley C.J.
      • Dahman B.
      • et al.
      Association of polypectomy techniques, endoscopist volume, and facility type with colonoscopy complications.
      • Bielawska B.
      • Day A.G.
      • Lieberman D.A.
      • et al.
      Risk factors for early colonoscopic perforation include non-gastroenterologist endoscopists: a multivariable analysis.
      • Blotière P.-O.
      • Weill A.
      • Ricordeau P.
      • et al.
      Perforations and haemorrhages after colonoscopy in 2010: a study based on comprehensive French health insurance data (SNIIRAM).
      • Pox C.P.
      • Altenhofen L.
      • Brenner H.
      • et al.
      Efficacy of a nationwide screening colonoscopy program for colorectal cancer.
      • Korman L.Y.
      • Overholt B.F.
      • Box T.
      • et al.
      Perforation during colonoscopy in endoscopic ambulatory surgical centers.
      Our study has limitations. Because the colonoscopy patients were selected in SASD based on exposure to colonoscopy, we could not construct a matched cohort without exposure to colonoscopy or another ambulatory procedure in SASD. However, we developed 3 comparator groups that we believe represent procedures with low systemic risk, and we compared our colonoscopy results to data on background event rates in the general population. Second, our Charlson-Deyo comorbidity scoring was based on comorbidities that occurred within 1 year of the colonoscopy in emergency, hospital, and ambulatory surgery encounters, but not routine outpatient encounters, which probably underestimates true comorbidities. Third, we did not study anesthesia use because it was not coded reliably in the database (CPT 00810 was used in only 0.12% of cases, HCPCS J3490 was not found at all, and there was no other information on medications such as propofol),
      • Cooper G.S.
      • Kou T.D.
      • Rex D.K.
      Complications following colonoscopy with anesthesia assistance: a population-based analysis.
      but the use of anesthesia for colonoscopy on the West Coast is nearly 7 times less frequent than in the Northeast,
      • Wernli K.J.
      • Brenner A.T.
      • Rutter C.M.
      • et al.
      Risks associated with anesthesia services during colonoscopy.
      and we focused solely on California. A recent study reported increased adjusted risks of complications with anesthesia services during colonoscopy.
      • Wernli K.J.
      • Brenner A.T.
      • Rutter C.M.
      • et al.
      Risks associated with anesthesia services during colonoscopy.
      Fourth, we were unable to ascertain endoscopist and facility characteristics.
      • Chukmaitov A.
      • Bradley C.J.
      • Dahman B.
      • et al.
      Association of polypectomy techniques, endoscopist volume, and facility type with colonoscopy complications.
      • Bielawska B.
      • Day A.G.
      • Lieberman D.A.
      • et al.
      Risk factors for early colonoscopic perforation include non-gastroenterologist endoscopists: a multivariable analysis.
      • Blotière P.-O.
      • Weill A.
      • Ricordeau P.
      • et al.
      Perforations and haemorrhages after colonoscopy in 2010: a study based on comprehensive French health insurance data (SNIIRAM).
      • Pox C.P.
      • Altenhofen L.
      • Brenner H.
      • et al.
      Efficacy of a nationwide screening colonoscopy program for colorectal cancer.
      Adenoma detection rates vary by colonoscopist,
      • Bretthauer M.
      • Kaminski M.F.
      • Løberg M.
      • et al.
      Population-based colonoscopy screening for colorectal cancer: a randomized clinical trial.
      • Chen S.C.
      • Rex D.K.
      Endoscopist can be more powerful than age and male gender in predicting adenoma detection at colonoscopy.
      and complication rates may also vary. Finally, we did not study splenic
      • Kamath A.S.
      • Iqbal C.W.
      • Sarr M.G.
      • et al.
      Colonoscopic splenic injuries: incidence and management.
      or acute kidney injury.
      • Layton J.B.
      • Klemmer P.J.
      • Christiansen C.F.
      • et al.
      Sodium phosphate does not increase risk for acute kidney injury after routine colonoscopy, compared with polyethylene glycol.
      In conclusion, our study provides a comprehensive assessment of post-colonoscopy GI and non-GI adverse event rates and provides a context for interpreting these rates. The threshold to tolerate complications is generally higher when procedures are performed to evaluate symptoms or deliver an intervention than when the purpose is preventive, as in screening/surveillance. Our findings suggest that S-Colo is associated with small but clinically significant risks of serious GI complications, and that these risks are higher when S-Colo includes biopsy/polypectomy, but polypectomy is precisely the intervention at screening/surveillance that may decrease CRC incidence. Most serious GI and non-GI complications led to hospitalization, and most serious and non-serious GI complications occurred within 14 days of colonoscopy. The rates of MI, stroke, and serious pulmonary event after S-Colo appear not to be higher than after low-risk comparator procedures, and the MI rates are similar to published background rates in the general population. However, stroke rates after S-Colo may be higher than the background population rates. These factors, as well as age and comorbidities, should be considered when weighing the benefits vs risks of S-Colo in specific clinical situations.

      Acknowledgments

      The authors thank Dr Subhas Banerjee for funding use of the SASD, SEDD, and SID databases for the study.

      Supplementary Material

      Supplementary Table 1CPT Codes for Colonoscopy Groups
      Diagnostic4438845378G0105G0121
      Biopsy4438945380
      Intervention443904439144392443934439444397453794538145382
      4538345384453854538645387
      Supplementary Table 2CPT Codes for Other Endoscopy Groups
      EGD432344323543236432394324143247432504325143254
      432554325643257432584326643270S2215
      Esophagoscopy431974319843200432014320243204432054321143212
      4321743288432284329943229
      Small bowel biopsy44361443764437744379
      Sigmoidoscopy45330G010445331453334533545338453394534545340
      453344533745332
      EGD, esophagogastroduodenoscopy.
      Supplementary Table 3CPT Codes for Control Groups
      Joint injections/aspirations and lithotripsy0229T0231T0228T0230T6448464479644836448050590206052060620610
      0215T0214T0217T0218T0213T0216T644906449364491644946449264495
      Minimally invasive joint surgeries298732988229876298702988329875298742988429867298682987729879
      2988129880298502988829889G0289298062980729819298052982229820
      298212982329825298262982829827S2114S23006472129848
      Cataract surgery6682066830668406685066852669206693066983669846694066982
      Supplementary Table 4ICD-9 Codes of Complications After Procedures
      GIColonic diverticulitis562.1x
      Small intestinal diverticulitis562.0x
      Noninfectious gastroenteritis558.9
      Lower GI bleeding569.3578.0578.1578.9998.1x
      Upper GI bleeding456.0456.20530.7530.82531.0x531.2x531.4x531.6x532.0x532.2x
      532.4x532.6x533.0x533.2x533.4x533.6x534.0x534.2x534.4x534.6x
      GI perforation569.83998.2E870.8E870.9
      Nausea787.0787.01787.02787.03536.2
      Abdominal pain789.0789.0x789.6x
      CardiacAcute MI410.x
      Coronary atherosclerosis411.x412413.x414.x429.2V45.81V45.8200.66
      00.66 is ICD-9 procedure code.
      Congestive heart failure398.91428428.0428.1428.2x428.3x428.4x428.9402.01402.11
      402.91404.11404.91404.03404.13404.93415.0
      Cardiac dysrhythmias427.x785.0785.1997.1V12.53
      Structural heart disease394.x395.x396.x397.x424.x996.71996.74
      CerebrovascularIschemic stroke346.0x346.6x433.x434436437.x438.x
      Hemorrhagic stroke430431432.0432.1432.9
      TIA435.xV12.54
      InfectionFever780.6780.62
      Bacteremia790.7
      PulmonaryPneumonia507507.0997.32997.39486
      Bronchitis466.x490491.21491.22494.1496491.0491.1491.2491.20
      491.8491.9492.0492.8494494.0
      GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a 00.66 is ICD-9 procedure code.
      Supplementary Table 5Comorbidities Based on Charlson Deyo Scoring
      ComplicationsComplications (specified)ICD-9
      Myocardial infarctionAcute myocardial infarction410–410.9x
      Old myocardial infarction412
      Congestive heart failureHeart failure428–428.9
      Peripheral vascular diseasePeripheral vascular disease unspecified443.9
      Aortic aneurysm and dissection441.0–441.9
      Gangrene785.4
      Blood vessel replaced by prosthesisV43.4
      Angiectomy
      Because procedure codes were available in CPT codes only, we mapped the ICD-9 procedure codes into CPT codes instead.
      38.48
      Because procedure codes were available in CPT codes only, we mapped the ICD-9 procedure codes into CPT codes instead.
      35141
      ICD-9 to CPT mapping for ICD-9 procedure.
      35142
      ICD-9 to CPT mapping for ICD-9 procedure.
      35151
      ICD-9 to CPT mapping for ICD-9 procedure.
      35152
      ICD-9 to CPT mapping for ICD-9 procedure.
      CerebrovascularCerebrovascular disease430–438.9
      DementiaDementia290–290.9
      Chronic pulmonary diseaseChronic obstructive pulmonary disease490–496
      Pneumoconioses500–505
      Chronic respiratory conditions due to fumes and vapors506.4
      Rheumatologic diseaseSystemic lupus erythematosus710.0
      Systemic sclerosis710.1
      Polymyositis710.4
      Adult rheumatoid arthritis714.0–714.2
      Rheumatoid lung714.81
      Polymyalgia rheumatica725
      Peptic ulcer diseaseGastric, duodenal, and gastrojejunal ulcers531–534.9
      Mild liver diseaseAlcoholic cirrhosis571.2
      Cirrhosis without mention of alcohol571.5
      Biliary cirrhosis571.6
      Chronic hepatitis571.4–571.49
      DiabetesDiabetes with or without acute metabolic disturbances250–250.3
      Diabetes with peripheral circulatory disorders250.7
      Diabetes with renal, ophthalmic, or neurologic manifestations250.4–250.6
      Hemiplegia or paraplegiaParaplegia344.1
      Hemiplegia342–342.9
      Renal diseaseChronic glomerulonephritis582–582.9
      Nephritis and nephropathy583–583.7
      Chronic renal failure585
      Renal failure, unspecified586
      Disorders resulting from impaired renal function588–588.9
      Any malignancyMalignant neoplasms140–172.9, 174–195.8
      Malignant neoplasms
      Leukemia and lymphoma200–208.9
      Liver diseaseHepatic coma, portal hypertension, other sequelae of chronic liver disease572.2–572.8
      Esophageal varices456.0–456.21
      Secondary malignant neoplasm of lymph nodes and other organs196–199.1
      AIDSAIDS042–044.98
      a Because procedure codes were available in CPT codes only, we mapped the ICD-9 procedure codes into CPT codes instead.
      b ICD-9 to CPT mapping for ICD-9 procedure.
      Supplementary Table 6Odds of Complications After Screening/Surveillance Colonoscopy Compared With Non-screening/Surveillance Colonoscopy
      Adjusted for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      ComplicationsDiagnosticBiopsy or intervention
      Odds ratios (95% CI)P valueOdds ratios (95% CI)P value
      Any complication0.65 (0.63–0.67)<.00010.66 (0.65–0.68)<.0001
      GI complication0.54 (0.52–0.57)<.00010.70 (0.68–0.72)<.0001
      Serious GI complication
      Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      0.44 (0.41–0.47)<.00010.70 (0.68–0.73)<.0001
       Lower GI bleeding0.44 (0.38–0.50)<.00010.82 (0.78–0.87)<.0001
       Upper GI bleeding0.54 (0.40–0.73)<.00010.72 (0.57–0.91).0053
       GI perforation0.70 (0.58–0.84).00020.61 (0.53–0.69)<.0001
       Colonic diverticulitis0.40 (0.36–0.43)<.00010.66 (0.62–0.70)<.0001
       Small bowel diverticulitis
      Model did not converge for small intestine diverticulitis.
      Non-serious GI complication
      Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      0.63 (0.60–0.67)<.00010.67 (0.64–0.70)<.0001
       Abdominal pain0.61 (0.57–0.66)<.00010.74 (0.70–0.78)<.0001
       Nausea0.72 (0.65–0.80)<.00010.67 (0.62–0.73)<.0001
       Gastroenteritis (noninfectious)0.54 (0.45–0.66)<.00010.37 (0.33–0.42)<.0001
      Non-GI0.69 (0.66–0.71)<.00010.65 (0.63–0.66)<.0001
      Cardiac0.70 (0.67–0.73)<.00010.71 (0.68–0.73)<.0001
       Acute MI0.79 (0.64–0.96).02100.71 (0.60–0.83)<.0001
       Coronary atherosclerosis0.67 (0.63–0.72)<.00010.72 (0.68–0.75)<.0001
       CHF0.67 (0.61–0.74)<.00010.73 (0.68–0.78)<.0001
       Cardiac dysrhythmia0.74 (0.69–0.79)<.00010.70 (0.67–0.73)<.0001
       Cardiac structural0.60 (0.52–0.69)<.00010.67 (0.61–0.74)<.0001
      Cerebrovascular0.74 (0.67–0.81)<.00010.75 (0.70–0.81)<.0001
       Ischemic stroke0.72 (0.63–0.84)<.00010.73 (0.66–0.81)<.0001
       Hemorrhagic stroke0.55 (0.34–0.87).01091.07 (0.69–1.66).7692
       TIA0.73 (0.64–0.82)<.00010.76 (0.70–0.83)<.0001
      Pulmonary0.64 (0.60–0.69)<.00010.71 (0.68–0.75)<.0001
       Pneumonia0.66 (0.59–0.74)<.00010.65 (0.59–0.71)<.0001
       Bronchitis0.65 (0.59–0.70)<.00010.76 (0.71–0.80)<.0001
      Infection
      Includes bacteremia and fever.
      0.68 (0.60–0.78)<.00010.52 (0.48–0.57)<.0001
       Bacteremia0.69 (0.45–1.05).08470.65 (0.47–0.89).0079
      Death0.65 (0.53–0.80)<.00010.57 (0.49–0.66)<.0001
      CHF, congestive heart failure; GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a Adjusted for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      b Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      c Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      d Includes bacteremia and fever.
      e Model did not converge for small intestine diverticulitis.
      Supplementary Table 7AComplications Within 30 Days of Comparator Procedure
      CharacteristicJoint injections, aspirations, lithotripsy (n = 324,114)Arthroscopies, carpal tunnel surgery (n = 853,170)Cataract surgery (n = 838,266)
      Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)
      Any complication9,087280 (275–286)12,791150 (147–153)17,671211 (208–214)
      GI3,542109 (106–113)4,38351 (50–53)3,85746 (45–47)
      Serious (GI)
      Includes GI perforation, upper and lower GI bleeding and diverticulitis.
      75723.4 (21.7–25.0)1,37316 (15–17)1,52318 (17–19)
       Lower GI bleeding2457.6 (6.6–8.5)85410.0 (9.3–10.7)5887.0 (6.4–7.6)
       Upper GI bleeding672.1 (1.6–2.6)921.1 (0.9–1.3)1591.9 (1.6–2.2)
       Perforation792.4 (1.9–3.0)420.5 (0.3–0.6)680.8 (0.6–1.0)
       Colonic diverticulitis40312.4 (11.2–13.6)4335.1 (4.6–5.6)8139.7 (9.0–10.4)
       Small diverticulitis80.2 (0.1–0.4)60.1 (0.0–0.1)170.2 (0.1–0.3)
      Non-serious (GI)
      Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      2,91389.9 (86.6–93.1)3,17337 (36–38)2,51930 (29–31)
       Abdominal pain1,90958.9 (56.3–61.5)1,34815.8 (15.0–16.6)1,24514.9 (14.0–15.7)
       Nausea1,16535.9 (33.9–38.0)1,80721.2 (20.2–22.2)1,15213.7 (12.9–14.5)
       Gastroenteritis (noninfectious)2337.2 (6.3–8.1)3454.0 (3.6–4.5)3804.5 (4.1–5.0)
      Non-GI6,228192 (187–197)7,81292 (90–94)15,460184 (182–187)
      Cardiac3,722115 (111–119)3,44340 (39–42)10,883130 (127–132)
       Acute MI1564.8 (4.1–5.6)1832.1 (1.8–2.5)6457.7 (7.1–8.3)
       Coronary atherosclerosis2,09564.6 (61.9–67.4)1,69419.9 (18.9–20.8)6,47577.2 (75.4–79.1)
       CHF77023.8 (22.1–25.4)6177.2 (6.7–7.8)3,68844.0 (42.6–45.4)
       Cardiac dysrhythmia1,75254.1 (51.5–56.6)1,74620.5 (19.5–21.4)5,10960.9 (59.3–62.6)
       Cardiac structural40112.4 (11.2–13.6)3384.0 (3.5–4.4)1,15913.8 (13.0–14.6)
      Cerebrovascular92928.7 (26.8–30.5)8209.6 (9.0–10.3)3,26238.9 (37.6–40.2)
       Ischemic stroke35210.9 (9.7–12.0)3223.8 (3.4–4.2)1,47117.5 (16.7–18.4)
       Hemorrhagic stroke260.8 (0.5–1.1)300.4 (0.2–0.5)1351.6 (1.3–1.9)
       TIA61118.9 (17.4–20.3)5416.3 (5.8–6.9)1,97723.6 (22.5–24.6)
      Pulmonary1,60750 (47–52)1,77320.8 (19.8–21.7)4,90358.5 (56.9–60.1)
       Pneumonia52216.1 (14.7–17.5)7168.4 (7.8–9.0)1,73020.6 (19.7–21.6)
       Bronchitis1,24438.4 (36.3–40.5)1,18813.9 (13.1–14.7)3,83145.7 (44.3–47.1)
      Infection
      Includes bacteremia and fever.
      37311.5 (10.3–12.7)2,04123.9 (22.9–25.0)5416.5 (5.9–7.0)
       Bacteremia672.1 (1.6–2.6)540.6 (0.5–0.8)1181.4 (1.2–1.7)
      Death1374.2 (3.5–4.9)911.1 (0.8–1.3)5256.3 (5.7–6.8)
      CHF, congestive heart failure; GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a Includes GI perforation, upper and lower GI bleeding and diverticulitis.
      b Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      c Includes bacteremia and fever.
      Supplementary Table 7BComplications Within 30 Days of Lithotripsy, Joint Aspirations, or Joint Injections
      CharacteristicLithotripsyJoint aspirationsJoint injections
      Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)Events (n)Event rate per 10,000 (95% CI)
      Any complication3,070446 (431–462)1,561282 (268–295)4,456223 (216–229)
      GI1,803262 (250–274)40673 (66–80)1,33367 (63–70)
      Serious (GI)
      Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      23934.7 (30.3–39.1)13023 (19–27)38819 (17–21)
       Lower GI bleeding9013.1 (10.4–15.8)519.2 (6.7–11.7)1045.2 (4.2–6.2)
       Upper GI bleeding223.2 (1.9–4.5)112.0 (0.8–3.2)341.7 (1.1–2.3)
       Perforation142.0 (1.0–3.1)61.1 (0.2–1.9)593.0 (2.2–3.7)
       Colonic diverticulitis12117.6 (14.5–20.7)6812.3 (9.4–15.2)21410.7 (9.3–12.1)
       Small bowel diverticulitis30.4 (-0.1–0.9)10.2 (-0.2–0.5)40.2 (0.0–0.4)
      Nonserious (GI)
      Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      1,609233.9 (222.6–245.2)29653 (47–59)1,00850 (47–54)
       Abdominal pain1,175170.8 (161.1–180.5)16329.4 (24.9–33.9)57128.6 (26.2–30.9)
       Nausea60587.9 (81.0–94.9)13223.8 (19.7–27.9)42821.4 (19.4–23.4)
       Gastroenteritis (noninfectious)395.7 (3.9–7.4)519.2 (6.7–11.7)1437.2 (6.0–8.3)
      Non-GI1,520221 (210–232)1,259227 (215–239)3,449173 (167–178)
      Cardiac822119 (111–128)747135 (125–144)2,153108 (103–112)
       Acute MI446.4 (4.5–8.3)325.8 (3.8–7.8)804.0 (3.1–4.9)
       Coronary atherosclerosis46567.6 (61.5–73.7)39571.2 (64.2–78.2)1,23561.8 (58.4–65.2)
       CHF12718.5 (15.3–21.7)18132.6 (27.9–37.4)46223.1 (21.0–25.2)
       Cardiac dysrhythmia36352.8 (47.4–58.2)37667.8 (61.0–74.6)1,01350.7 (47.6–53.8)
       Cardiac structural7711.2 (8.7–13.7)7213.0 (10.0–16.0)25212.6 (11.1–14.2)
      Cerebrovascular21831.7 (27.5–35.9)17431.4 (26.7–36.0)53726.9 (24.6–29.1)
       Ischemic stroke8912.9 (10.3–15.6)6411.5 (8.7–14.4)19910.0 (8.6–11.3)
       Hemorrhagic stroke71.0 (0.3–1.8)50.9 (0.1–1.7)140.7 (0.3–1.1)
       TIA13920.2 (16.8–23.6)11821.3 (17.4–25.1)35417.7 (15.9–19.6)
      Pulmonary28642 (37–46)30755.4 (49.2–61.5)1,01450.7 (47.6–53.8)
       Pneumonia9614.0 (11.2–16.7)9216.6 (13.2–20.0)33416.7 (14.9–18.5)
       Bronchitis21531.3 (27.1–35.4)24644.4 (38.8–49.9)78339.2 (36.4–41.9)
      Infection
      Includes bacteremia and fever.
      16223.5 (19.9–27.2)9517.1 (13.7–20.6)1165.8 (4.7–6.9)
       Bacteremia324.7 (3.0–6.3)122.2 (0.9–3.4)231.2 (0.7–1.6)
      Death304.4 (2.8–5.9)336.0 (3.9–8.0)743.7 (2.9–4.5)
      CHF, congestive heart failure; GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      b Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      c Includes bacteremia and fever.
      Supplementary Table 8Odds of Complications After Screening/Surveillance Colonoscopy Compared With Lithotripsy, Joint Aspirations, or Joint Injections
      Adjusted for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      ComplicationsOdds ratios (95% CI)
      LithotripsyJoint aspirationsJoint injections
      Any complication0.39 (0.37–0.40)0.72 (0.68–0.76)0.79 (0.77–0.82)
      GI complication0.36 (0.34–0.38)1.33 (1.20–1.47)1.36 (1.28–1.44)
      Serious GI complication
      Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      1.50 (1.31–1.71)2.31 (1.94–2.76)2.57 (2.31–2.85)
       Lower GI hemorrhage2.06 (1.67–2.55)2.93 (2.22–3.87)4.89 (4.01–5.97)
       GI perforation2.36 (1.38–4.02)5.10 (2.28–11.41)1.76 (1.33–2.32)
      Nonserious GI complication
      Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      0.19 (0.18–0.20)0.85 (0.76–0.96)0.84 (0.78–0.90)
      Non-GI0.52 (0.49–0.55)0.62 (0.59–0.66)0.69 (0.66–0.71)
      Cardiac0.55 (0.51–0.60)0.63 (0.58–0.68)0.64 (0.61–0.67)
       Acute MI0.50 (0.36–0.69)0.68 (0.47–0.98)0.76 (0.60–0.97)
       CHF0.84 (0.69–1.01)0.64 (0.55–0.75)0.68 (0.61–0.76)
       Cardiac dysrhythmia0.63 (0.57–0.71)0.64 (0.58–0.72)0.73 (0.68–0.79)
      Cerebrovascular0.47 (0.41–0.54)0.70 (0.59–0.82)0.64 (0.58–0.70)
       Ischemic stroke0.49 (0.39–0.61)0.78 (0.60–1.01)0.71 (0.61–0.83)
       Hemorrhagic stroke0.46 (0.21–1.00)0.90 (0.33–2.47)0.77 (0.43–1.38)
       TIA0.46 (0.39–0.55)0.65 (0.54–0.79)0.61 (0.54–0.69)
      Pulmonary0.73 (0.64–0.83)0.69 (0.61–0.78)0.60 (0.56–0.65)
       Pneumonia0.70 (0.57–0.87)0.75 (0.60–0.93)0.61 (0.54–0.69)
       Bronchitis0.75 (0.65–0.86)0.68 (0.59–0.78)0.60 (0.56–0.66)
      Infection
      Includes bacteremia and fever.
      0.45 (0.38–0.53)0.58 (0.47–0.72)1.63 (1.34–1.99)
       Bacteremia0.18 (0.12–0.27)0.44 (0.23–0.82)0.63 (0.40–1.01)
      Death0.66 (0.45–0.96)0.68 (0.47–0.98)0.85 (0.66–1.10)
      CHF, congestive heart failure; GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a Adjusted for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      b Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      c Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      d Includes bacteremia and fever.
      Supplementary Table 9Odds of Complications After Non-screening/Surveillance Colonoscopy Compared With Non-colonoscopy Procedures Presumed to have Low Systemic Risk
      Adjusted for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      ComplicationsOdds ratios (95% CI)
      Joint injections, aspirations, lithotripsyArthroscopies, carpal tunnel surgeryCataract surgery
      Any complication0.99 (0.96–1.02)1.54 (1.50–1.57)2.15 (2.11–2.20)
      GI complication1.31 (1.26–1.36)2.63 (2.54–2.73)4.10 (3.94–4.27)
      Serious GI complication
      Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      3.37 (3.12–3.64)4.44 (4.17–4.72)8.42 (7.93–8.95)
       Lower GI bleeding3.82 (3.35–4.37)2.61 (2.39–2.84)7.21 (6.54–7.95)
       GI perforation3.27 (2.58–4.16)12.01 (8.71–16.57)15.53 (11.93–20.21)
      Non-serious GI complication
      Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      0.81 (0.77–0.85)1.90 (1.81–1.99)2.30 (2.18–2.43)
      Non-GI0.95 (0.92–0.98)1.48 (1.44–1.52)1.91 (1.86–1.96)
      Cardiac0.90 (0.87–0.94)1.71 (1.64–1.78)1.78 (1.73–1.84)
       Acute MI0.95 (0.78–1.15)1.37 (1.13–1.64)1.22 (1.07–1.40)
       CHF1.03 (0.95–1.12)1.95 (1.77–2.14)1.41 (1.33–1.50)
       Cardiac dysrhythmia1.01 (0.95–1.06)1.77 (1.67–1.87)1.93 (1.85–2.02)
      Cerebrovascular0.82 (0.76–0.89)1.47 (1.35–1.60)1.41 (1.33–1.50)
       Ischemic stroke0.93 (0.82–1.05)1.61 (1.41–1.84)1.40 (1.28–1.54)
       Hemorrhagic stroke0.98 (0.62–1.55)1.34 (0.86–2.08)0.93 (0.68–1.28)
       TIA0.79 (0.71–0.87)1.39 (1.25–1.54)1.45 (1.34–1.56)
      Pulmonary0.95 (0.89–1.00)1.55 (1.46–1.64)1.53 (1.47–1.61)
       Pneumonia1.00 (0.91–1.11)1.31 (1.19–1.44)1.47 (1.36–1.59)
       Bronchitis0.92 (0.86–0.99)1.71 (1.59–1.84)1.50 (1.42–1.58)
      Infection
      Includes bacteremia and fever.
      1.36 (1.21–1.53)0.57 (0.53–0.62)2.98 (2.67–3.32)
       Bacteremia0.53 (0.39–0.73)1.19 (0.84–1.68)1.35 (1.00–1.81)
      Death1.32 (1.09–1.60)2.93 (2.32–3.70)1.88 (1.64–2.15)
      CHF, congestive heart failure; GI, gastrointestinal; MI, myocardial infarction; TIA, transient ischemic attack.
      a Adjusted for age, gender, race, insurance, median income, other endoscopies within 30 days of colonoscopy, Charlson-Deyo comorbidity score, patient location, and number of chronic conditions.
      b Includes GI perforation, upper and lower GI bleeding, and diverticulitis.
      c Includes nausea, abdominal pain, and noninfectious gastroenteritis.
      d Includes bacteremia and fever.

      References

        • Peery A.F.
        • Dellon E.S.
        • Lund J.
        • et al.
        Burden of gastrointestinal disease in the United States: 2012 update.
        Gastroenterology. 2012; 143 (e1–e3): 1179-1187
        • Seeff L.C.
        • Richards T.B.
        • Shapiro J.A.
        • et al.
        How many endoscopies are performed for colorectal cancer screening? Results from CDC’s survey of endoscopic capacity.
        Gastroenterology. 2004; 127: 1670-1677
        • Bacchus C.M.
        • Dunfield L.
        • et al.
        • Canadian Task Force on Preventive Health Care
        Recommendations on screening for colorectal cancer in primary care.
        CMAJ. 2016; 188: 340-348
        • Zauber A.G.
        • Knudsen A.B.
        • Rutter C.M.
        • et al.
        Evaluating the benefits and harms of colorectal cancer screening strategies: a collaborative modeling approach.
        AHRQ. 2015; 14: 1-122
        • Lin J.
        • Piper M.
        • Perdue L.
        • et al.
        Screening for colorectal cancer: an updated systematic review for the USPSTF.
        AHRQ. 2015; 14: 1-240
      1. Agency for Healthcare Research and Quality. HCUP quality control procedures. 2017. Available at: https://www.hcup-us.ahrq.gov/db/quality.pdf. (Accessed May 15, 2017).

      2. Agency for Healthcare Research and Quality. Introduction to the HCUP State Ambulatory Surgery and Services Databases (SASD). 2017. Available at: https://www.hcup-us.ahrq.gov/db/state/sasddist/Introduction_to_SASD.pdf. (Accessed May 15, 2017).

      3. Agency for Healthcare Research and Quality. Introduction to the HCUP State Emergency Department Databases (SEDD). 2017. Available at: https://www.hcup-us.ahrq.gov/db/state/sedddist/Introduction_to_SEDD.pdf. (Accessed May 15, 2017).

      4. Agency for Healthcare Research and Quality. Introduction to the HCUP State Inpatient Databases (SID). 2017. Available at: https://www.hcup-us.ahrq.gov/db/state/siddist/Introduction_to_SID.pdf. (Accessed May 15, 2017).

      5. Agency for Healthcare Research and Quality. HCUP supplemental variables for revisit analyses. Available at: https://hcup-us.ahrq.gov/toolssoftware/revisit/revisit.jsp. (Accessed May 15, 2017).

        • Ladabaum U.
        • Levin Z.
        • Mannalithara A.
        • et al.
        Colorectal testing utilization and payments in a large cohort of commercially insured US adults.
        Am J Gastroenterol. 2014; 109: 1513-1525
        • Ko C.W.
        • Dominitz J.A.
        Complications of colonoscopy: magnitude and management.
        Gastrointest Endosc Clin N Am. 2010; 20: 659-671
        • Arora G.
        • Mannalithara A.
        • Singh G.
        • et al.
        Risk of perforation from a colonoscopy in adults: a large population-based study.
        Gastrointest Endosc. 2009; 69: 654-664
        • Sporn E.
        • Petroski G.F.
        • Mancini G.J.
        • et al.
        Laparoscopic appendectomy–is it worth the cost? Trend analysis in the US from 2000 to 2005.
        J Am Coll Surg. 2009; 208: 179-185.e2
        • Kreder H.J.
        • Deyo R.A.
        • Koepsell T.
        • et al.
        Relationship between the volume of total hip replacements performed by providers and the rates of postoperative complications in the state of Washington.
        J Bone Joint Surg Am. 1997; 79: 485-494
        • Deyo R.A.
        • Cherkin D.C.
        • Ciol M.A.
        Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases.
        J Clin Epidemiol. 1992; 45: 613-619
        • Reumkens A.
        • Rondagh E.J.A.
        • Bakker C.M.
        • et al.
        Post-colonoscopy complications: a systematic review, time trends, and meta-analysis of population-based studies.
        Am J Gastroenterol. 2016; 111: 1092-1101
        • Ko C.W.
        • Riffle S.
        • Michaels L.
        • et al.
        Serious complications within 30 days of screening and surveillance colonoscopy are uncommon.
        Clin Gastroenterol Hepatol. 2010; 8: 166-173
        • Warren J.L.
        Adverse events after outpatient colonoscopy in the Medicare population.
        Ann Intern Med. 2009; 150 (849–810)
        • Rabeneck L.
        • Paszat L.F.
        • Hilsden R.J.
        • et al.
        Bleeding and perforation after outpatient colonoscopy and their risk factors in usual clinical practice.
        Gastroenterology. 2008; 135: 1899-1906.e1
        • Levin T.R.
        • Zhao W.
        • Conell C.
        • et al.
        Complications of colonoscopy in an integrated health care delivery system.
        Ann Intern Med. 2006; 145: 880-889
        • Basson M.D.
        • Etter L.
        • Panzini L.A.
        Rates of colonoscopic perforation in current practice.
        Gastroenterology. 1998; 114: 1115
        • Kang H.Y.
        • Kang H.W.
        • Kim S.G.
        • et al.
        Incidence and management of colonoscopic perforations in Korea.
        Digestion. 2008; 78: 218-223
        • Polter D.E.
        Risk of colon perforation during colonoscopy at Baylor University Medical Center.
        Proc (Bayl Univ Med Cent). 2015; 28: 3-6
        • Lippert E.
        • Herfarth H.H.
        • Grunert N.
        • et al.
        Gastrointestinal endoscopy in patients aged 75 years and older: risks, complications, and findings—a retrospective study.
        Int J Colorectal Dis. 2015; 30: 363-366
        • Kahi C.J.
        • Azzouz F.
        • Juliar B.E.
        • et al.
        Survival of elderly persons undergoing colonoscopy: implications for colorectal cancer screening and surveillance.
        Gastrointest Endosc. 2007; 66: 544-550
        • Rathgaber S.W.
        • Wick T.M.
        Colonoscopy completion and complication rates in a community gastroenterology practice.
        Gastrointest Endosc. 2006; 64: 556-562
        • Zubarik R.
        • Fleischer D.
        • Mastropietro C.
        • et al.
        Prospective analysis of complications 30 days after outpatient colonoscopy.
        Gastrointest Endosc. 1999; 50: 1-7
        • Lüning T.H.
        • Keemers-Gels M.E.
        • Barendregt W.B.
        • et al.
        Colonoscopic perforations: a review of 30,366 patients.
        Surg Endosc. 2007; 21: 994-997
        • Eckardt V.F.
        • Kanzler G.
        • Schmitt T.
        • et al.
        Complications and adverse effects of colonoscopy with selective sedation.
        Gastrointest Endosc. 1999; 49: 560-565
        • Sieg A.
        • Hachmoeller-Eisenbach U.
        • Eisenbach T.
        Prospective evaluation of complications in outpatient GI endoscopy: a survey among German gastroenterologists.
        Gastrointest Endosc. 2001; 53: 620-627
        • Nelson D.B.
        • McQuaid K.R.
        • Bond J.H.
        Procedural success and complications of large-scale screening colonoscopy.
        Gastrointest Endosc. 2002; 55: 307-314
        • Viiala C.H.
        • Zimmerman M.
        • Cullen D.J.E.
        • et al.
        Complication rates of colonoscopy in an Australian teaching hospital environment.
        Intern Med J. 2003; 33: 355-359
        • Geenen J.E.
        • Schmitt M.G.
        • Wu W.C.
        • et al.
        Major complications of coloscopy: bleeding and perforation.
        Am J Dig Dis. 1975; 20: 231-235
        • Farley D.R.
        • Bannon M.P.
        • Zietlow S.P.
        • et al.
        Management of colonoscopic perforations.
        Mayo Clinic Proc. 1997; 72: 729-733
        • Anderson M.L.
        • Pasha T.M.
        • Leighton J.A.
        Endoscopic perforation of the colon: lessons from a 10-year study.
        Am J Gastroenterol. 2000; 95: 3418-3422
        • Chukmaitov A.
        • Bradley C.J.
        • Dahman B.
        • et al.
        Association of polypectomy techniques, endoscopist volume, and facility type with colonoscopy complications.
        Gastrointest Endosc. 2013; 77: 436-446
        • Bielawska B.
        • Day A.G.
        • Lieberman D.A.
        • et al.
        Risk factors for early colonoscopic perforation include non-gastroenterologist endoscopists: a multivariable analysis.
        Clin Gastroenterol Hepatol. 2014; 12: 85-92
        • Singh H.
        • Penfold R.B.
        • DeCoster C.
        • et al.
        Colonoscopy and its complications across a Canadian regional health authority.
        Gastrointest Endosc. 2009; 69: 665-671
        • Adeyemo A.
        • Bannazadeh M.
        • Riggs T.
        • et al.
        Does sedation type affect colonoscopy perforation rates?.
        Dis Colon Rectum. 2014; 57: 110-114
        • Blotière P.-O.
        • Weill A.
        • Ricordeau P.
        • et al.
        Perforations and haemorrhages after colonoscopy in 2010: a study based on comprehensive French health insurance data (SNIIRAM).
        Clin Res Hepatol Gastroenterol. 2014; 38: 112-117
        • Cooper G.S.
        • Kou T.D.
        • Rex D.K.
        Complications following colonoscopy with anesthesia assistance: a population-based analysis.
        JAMA Intern Med. 2013; 173: 551-556
        • Dominitz J.A.
        • Baldwin L.M.
        • Green P.
        • et al.
        Regional variation in anesthesia assistance during outpatient colonoscopy is not associated with differences in polyp detection or complication rates.
        Gastroenterology. 2013; 144: 298-306
        • Hamdani U.
        Risk factors for colonoscopic perforation: a population-based study of 80118 cases.
        World J Gastroenterol. 2013; 19 (3596–3596)
        • Stock C.
        • Ihle P.
        • Sieg A.
        • et al.
        Adverse events requiring hospitalization within 30 days after outpatient screening and nonscreening colonoscopies.
        Gastrointest Endosc. 2013; 77: 419-429
        • Quintero E.
        • Castells A.
        • Bujanda L.
        • et al.
        Colonoscopy versus fecal immunochemical testing in colorectal-cancer screening.
        N Engl J Med. 2012; 366: 697-706
        • Mozaffarian D.
        • Benjamin E.J.
        • Go A.S.
        • et al.
        Heart disease and stroke statistics–2015 update: a report from the American Heart Association.
        Circulation. 2015; 131: e29-e322
        • Rutter C.M.
        • Johnson E.
        • Miglioretti D.L.
        • et al.
        Adverse events after screening and follow-up colonoscopy.
        Cancer Causes Control. 2011; 23: 289-296
        • Ranasinghe I.
        • Parzynski C.S.
        • Searfoss R.
        • et al.
        Differences in colonoscopy quality among facilities: development of a post-colonoscopy risk-standardized rate of unplanned hospital visits.
        Gastroenterology. 2016; 150: 103-113
        • Holt C.L.
        • Shipp M.
        • Eloubeidi M.
        • et al.
        Use of focus group data to develop recommendations for demographically segmented colorectal cancer educational strategies.
        Health Educ Res. 2009; 24: 876-889
        • Wilkins T.
        • Gillies R.A.
        • Harbuck S.
        • et al.
        Racial disparities and barriers to colorectal cancer screening in rural areas.
        J Am Board Fam Med. 2012; 25: 308-317
        • Bromley E.G.
        • May F.P.
        • Federer L.
        • et al.
        Explaining persistent under-use of colonoscopic cancer screening in African Americans: a systematic review.
        Prev Med. 2015; 71: 40-48
        • James A.S.
        • Daley C.M.
        • Greiner K.A.
        Knowledge and attitudes about colon cancer screening among African Americans.
        Am J Health Behav. 2011; 35: 393-401
        • Siegel R.
        • DeSantis C.
        • Jemal A.
        Colorectal cancer statistics, 2014.
        CA Cancer J Clin. 2014; 64: 104-117
        • Almario C.V.
        • May F.P.
        • Ponce N.A.
        • et al.
        Racial and ethnic disparities in colonoscopic examination of individuals with a family history of colorectal cancer.
        Clin Gastroenterol Hepatol. 2015; 13: 1487-1495
        • Laiyemo A.O.
        • Thompson N.
        • Williams C.D.
        • et al.
        Race and colorectal cancer screening compliance among persons with a family history of cancer.
        World J Gastrointest Endosc. 2015; 7: 1300-1305
        • Decker K.M.
        • Demers A.A.
        • Nugent Z.
        • et al.
        Longitudinal rates of colon cancer screening use in Winnipeg, Canada: the experience of a universal health-care system with an organized colon screening program.
        Am J Gastroenterol. 2015; 110: 1640-1646
        • O'Malley A.S.
        • Forrest C.B.
        • Feng S.
        • et al.
        Disparities despite coverage: gaps in colorectal cancer screening among Medicare beneficiaries.
        Arch Intern Med. 2005; 165: 2129-2135
        • May F.P.
        • Bromley E.G.
        • Reid M.W.
        • et al.
        Low uptake of colorectal cancer screening among African Americans in an integrated Veterans Affairs health care network.
        Gastrointest Endosc. 2014; 80: 291-298
        • Haddad J.D.
        • You D.M.
        Colorectal cancer screening and race in an equal access medical system.
        J Community Health. 2016; 41: 78-81
        • Gatto N.M.
        • Frucht H.
        • Sundararajan V.
        • et al.
        Risk of perforation after colonoscopy and sigmoidoscopy: a population-based study.
        J Natl Cancer Inst. 2003; 95: 230-236
        • Zafar H.M.
        • Harhay M.O.
        • Yang J.
        • et al.
        Adverse events following computed tomographic colonography compared to optical colonoscopy in the elderly.
        Prev Med Rep. 2014; 1: 3-8
        • Hsieh T.K.
        • Hung L.
        • Kang F.C.
        • et al.
        Anesthesia does not increase the rate of bowel perforation during colonoscopy: a retrospective study.
        Acta Anaesthesiol Taiwan. 2009; 47: 162-166
        • Kamath A.S.
        • Iqbal C.W.
        • Sarr M.G.
        • et al.
        Colonoscopic splenic injuries: incidence and management.
        J Gastrointest Surg. 2009; 13: 2136-2140
        • Tam M.S.
        • Abbas M.A.
        Perforation following colorectal endoscopy: what happens beyond the endoscopy suite?.
        Perm J. 2013; 17: 17-21
        • Pox C.P.
        • Altenhofen L.
        • Brenner H.
        • et al.
        Efficacy of a nationwide screening colonoscopy program for colorectal cancer.
        Gastroenterology. 2012; 142: 1460-1467.e2
        • Korman L.Y.
        • Overholt B.F.
        • Box T.
        • et al.
        Perforation during colonoscopy in endoscopic ambulatory surgical centers.
        Gastrointest Endosc. 2003; 58: 554-557
        • Wernli K.J.
        • Brenner A.T.
        • Rutter C.M.
        • et al.
        Risks associated with anesthesia services during colonoscopy.
        Gastroenterology. 2016; 150: 888-894
        • Bretthauer M.
        • Kaminski M.F.
        • Løberg M.
        • et al.
        Population-based colonoscopy screening for colorectal cancer: a randomized clinical trial.
        JAMA Intern Med. 2016; 176: 894-902
        • Chen S.C.
        • Rex D.K.
        Endoscopist can be more powerful than age and male gender in predicting adenoma detection at colonoscopy.
        Am J Gastroenterol. 2007; 102: 856-861
        • Layton J.B.
        • Klemmer P.J.
        • Christiansen C.F.
        • et al.
        Sodium phosphate does not increase risk for acute kidney injury after routine colonoscopy, compared with polyethylene glycol.
        Clin Gastroenterol Hepatol. 2014; 12: 1514-1521.e3